Drone-Enabled Practices in Modern Warehouse Management: A Comprehensive Review

Highlights

What are the main findings?

• This study builds a clear taxonomy of warehouse-drone research (tasks, platforms, single vs multi-drone systems, simulation vs real-world validation, AI/IoT stack).
• The main application areas (inventory and mapping, inspection/maintenance, intralogistics transport/support, security/surveillance, picking support, layout optimization) are mapped.

What are the implications of the main findings?

• The maturity gap is revealed: research is still single-UAV and simulation-heavy, with few fully autonomous real-warehouse deployments.
• The key gaps cited are highlighted: multi-drone coordination/safety remains unresolved, and dynamic warehouse environments are under-studied.

Abstract

The advent of drone technology has led to groundbreaking advancements across various industries, including warehousing operations. In recent years, warehouse drones have garnered significant attention due to their potential to revolutionize traditional inventory management and order fulfillment processes. This paper presents a comprehensive review that synthesizes findings from more than 120 research papers on drone-enabled practices in warehouses. The review systematically considers multiple parameters, including drone function (inventory counting, mapping, surveillance, inspection, and intralogistics support), robot platforms used (UAV, UAV-AGV), deployment architecture (single and multi-drone system), validation approach (real-time and simulation), technology and methodology used (modern electronic devices, AI, and IOT), and environmental context (dynamic and static). Furthermore, the paper explores the diverse applications of warehouse drones in inventory management, maintenance and inspection, picking and packaging, goods transportation, security and surveillance, and warehouse layout optimization. The review highlights that most studies still rely on single-UAV systems tested mainly in simulations, with only a few real-time demonstrations of fully autonomous performance inside real warehouses. Although multi-drone approaches are emerging to improve scalability, they continue to struggle with coordination and safety. Research remains largely focused on static environments, with dynamic warehouse conditions receiving far less attention despite their practical importance. The findings of the review are presented with the tabulated results and a comparative table to provide a better understanding of the review work, which helps to identify the existing literature gap. The review presents its findings through clear tables and comparisons, making it easier to understand existing studies and pinpoint the gaps in the current literature.

1. Introduction

Warehouse Management (WM) [1] deals with overseeing, optimizing, and controlling the various operations and functions within a warehouse related to handling goods, from raw materials to finished goods, and storing them. It also involves several activities, such as optimizing the use of manpower, space, and technology to manage inventory and fulfill orders in the most optimal way. The goal of WM is to increase the efficiency of operations, minimize the cost involved in material handling, and ensure the timely delivery of goods to customers. In WM, there are various key components, such as inventory management (IM), order fulfillment, space utilization, technology utilization, quality control, safety and compliance, etc. (as shown in Figure 1).

Among all the key components in a WM system, inventory management [2] plays an important role in maintaining the stock level as per the manufacturing requirements and avoiding stockouts and overstock situations. IM ensures that the right product is available at the right time with the right quantity to fulfill the needs of the customer and market. The real-time monitoring of stock and stock movement is the key function of tracking stocks in IM. Stock replenishment in IM helps to restock materials based on demand. Stocking and auditing systems conduct regular stock audits to verify the inventory’s accuracy and minimize differences. Order fulfillment in the WM system processes the picking, packing, and shipping of the order to the customer under time-bound conditions. Picking strategies are required to optimize order retrieval, reduce labor costs, improve warehouse efficiency, etc. Implementing the batch picking (picking multiple orders), zone picking (retrieving items from their designated zones), and wave picking (time-based picking method where orders are grouped into waves based on priority, shipping schedules, or order characteristics) strategies can ensure increased order fulfillment speed, reduced picker congestion, improved accuracy, and reduced bottlenecks. Packing and labeling ensure the product is safely packed to protect it from any unwanted damage until it reaches the customer’s hand, and labeling provides a way to track and identify the product using barcodes, radio frequency identifications, etc. Shipping and delivery optimization in WM ensures that the product will reach the customer on time without any transit delay at a minimum cost. Its performance strongly depends on route optimization, real-time tracking, and automated carrier selection. To reduce the operational cost and to optimize storage capacity, space utilization is an important aspect of WM.

Space utilization consists of designing the warehouse layout to minimize the movement of the material and maximize storage capacity by employing vertical storage and pallet racking. Slotting optimization in WM categorizes material based on demand frequency. Materials that have high demand are kept in easily accessible areas. Cross-docking in space utilization refers to reducing storage time by directly transferring goods from inbound docks to outbound docks. Cross-docking reduces the cost of storage and the cost of handling material, improves supply chain management, and provides faster order fulfillment and better inventory turnover.

In modern warehouses [3], the application of technology is most commonly used to perform various tasks automatically. Modern equipment in warehouse management systems tracks inventory, manages orders, and optimizes various operations in warehouses; robotics and automation perform material handling operations without human intervention (examples are automated guided vehicles, drones, etc.), and IoT and AI systems provide real-time monitoring to keep an eye on predictive maintenance, asset tracking, and workflow automation in warehouses. These technologies help in facilitating a centralized software platform to track inventory, manage orders, reduce errors, perform repetitive tasks, conduct fast and accurate stock audits, and optimize warehouse operation, stock replenishment, order picking, shipment tracking, real-time visibility into inventory levels, enterprise resource planning, transportation management systems, and much more. Quality control in WM ensures the inspection of goods through quality checks before delivering them to the customer. It consists of a quality check for defects or damage in the product before dispatch. Return and reverse logistics management in WM is the process of handling returned material and optimizing the processes of resale, recycling, or disposal. This process in the warehouse management system enhances customer satisfaction, reduces waste, and enhances overall supply chain processes. The warehouse environment consists of continuous activities that involve heavy machinery, equipment, and humans; therefore, safety is paramount in WM. Workshop safety measures involve CCTV monitoring, ergonomic workstations, and safety gear to protect workers. Regulatory compliance with international standards such as the ISO (International Organization for Standardization), OSHA (Occupational Safety and Health Administration), and environmental regulations ensures legal and operational safety. Risk management and security ensure the protection of inventory from theft, cyber threats, and environmental hazards.

Traditional WMS [4] was completely manual and labor-intensive, and record-keeping for inventory was carried out using paper-based logs. The use of such systems in tracking inventory was cumbersome, prone to errors, and not reliable. It led to inefficiencies and higher operational costs. Due to a lack of advanced technologies in the traditional system, material visibility was limited, and hence, it was very difficult to monitor inventory levels in real-time. Issues such as miscounts and misplaced items in stock management ultimately affected order accuracy and resulted in order fulfillment being slow and inefficient. The main cause for this inefficient performance was the adoption of manual practices to locate and pick the inventory, which led to an increase in processing time and caused delays in shipment. In traditional workshops, the storage space remained largely static, and space utilization was not optimized, resulting in inefficient storage and retrieval processes. Such warehouses are not upgraded with supply chain partners, making coordination difficult and delaying communication across the logistics network. The maintenance of such warehouses is reactive rather than predictive, resulting in unexpected downtime and reduced productivity.

Hence, in order to overcome these drawbacks and improve performance, there is a need to modernize traditional warehouses by integrating state-of-the-art advanced technologies consisting of upgraded hardware and software [5]. Intelligent technology, consisting of robotics, automation, IoT, and AI, may be the game-changer in addressing current warehouse management challenges. These modern techniques leverage digital and cloud-based management systems to improve efficiency, accuracy, and real-time decision-making. They play a crucial role in streamlining the material handling system with reduced manpower and operational errors. Their real-time inventory tracking features, based on barcode scanners, IoT, AI, and RFID, ensure the precise monitoring of materials with a reduced risk of stockouts or overstocking. An automated picking system not only improves the order fulfillment speed and accuracy but also reduces the processing time, assuring customer satisfaction. Modern tools with AI optimization and decision-making provide optimal use of space through dynamic slotting and cross-docking strategies. This facilitates efficient storage and faster product movement. ERP, SCM, and e-commerce platforms are integral parts of the modern system that enable the smooth coordination of work across all operational levels. Sensor-based technology plays an important role in preventing sudden failures of warehouse operations and, hence, ensuring uninterrupted warehouse operations over time. These features of modern warehouse practices enhance warehouse efficiency, reduce operational costs, and improve overall supply chain performance. The key components of WM systems are shown in Figure 2.

Nowadays, the incredible transformation of drone technology [6] has ignited a paradigm shift in all industries, and one area that has witnessed significant transformation is warehouse management. Drone technology has evolved as a unique solution to warehouse management in various industries to address inventory management, stocktaking, order fulfillment, last-mile delivery, security, maintenance, inspection, mapping, layout optimization, etc. These drones provide a wide range of capabilities, from real-world inventory tracking to autonomous order fulfillment, ensuring enhanced efficiency, accuracy, cost-effectiveness, and user-friendliness [7].

However, research on UAV-enabled warehousing remains distributed across applications (e.g., inventory, inspection, security, picking support, and layout analysis), often emphasizing different sensing modalities, autonomy stacks, and evaluation metrics. This fragmentation creates a knowledge gap: stakeholders lack a consolidated, application-oriented synthesis that clearly contrasts drone-enabled warehouse management with traditional practices while identifying consistent performance drivers, limitations, and open research opportunities. Accordingly, this review aims to consolidate and structure the literature on warehouse drones, highlight their operational value and constraints, and map emerging directions that could shape the next generation of intelligent warehousing.

The remainder of this paper is organized as follows: Section 2 describes the review methodology; Section 3 introduces drone technology in warehouses; Section 4 details major warehouse-drone applications; Section 5 presents the results and discussion; and Section 6 concludes with future scope.

2. Review Methodology

The main aim of the review is to obtain insights into the role of drone technology in WM. It also provides in-depth detailing of the work that has been carried out so far in the field of supply chain management. The review paper highlights the effectiveness and advantages of drone integration in these applications. It also aims to conduct a comparative analysis of traditional warehouse management systems with drone-based warehouse management systems. It aims to give a fair assessment of the benefits and limitations of using drones in warehouses. This paper identifies the literature and research opportunities in warehouse management using drones. It also offers a glimpse into the potential future development of warehouse drones. The study imagines a future in which drone technology is essential to improving supply chain management and transforming warehousing operations, considering developments in artificial intelligence, sensor technology, and battery life. The overall goals of this in-depth analysis are to give readers a complete grasp of the state of warehouse drones today, their uses, and the potential revolution they could bring to warehousing operations. The proposed paper cites more than 120 research papers to conduct an in-depth survey.

The systematic review process filtered 453 initially downloaded UAV-warehouse research papers through duplicate removal, title-based and abstract/conclusion screening, and full-text assessment, finally resulting in 118 included studies classified into five application domains: inventory management, maintenance, picking/packing, security/surveillance, and warehouse layout optimization. The review mechanism of the paper is shown in Figure 3. The journal distribution analysis in Figure 4 shows research concentrated in Sensors, Drones, IEEE Access, and other applied automation outlets, indicating strong alignment with sensing, autonomy, and remote-sensing technologies. The keyword frequency shown in Figure 5 further confirms dominant research themes around UAV platforms, RFID-based identification, path planning, computer vision, localization, and multi-robot execution, while the lower weighting of SLAM, UGV, optimization, LiDAR, and blockchain suggests emerging interest in advanced autonomy and secure data exchange. The citation-based impact metrics shown in

Table 1. Research impact metrics summary.

Metric	Value	Analysis
Total Papers Analyzed	128	Comprehensive literature review
Total Citations (approx.)	~5424	Based on reported counts
Average Citations/Paper	~39	Skewed by highly cited papers
Median Citations/Paper	~12	More representative of a typical paper
Papers with 100+ citations	5	3.6% of papers
Papers with 50–99 citations	8	5.7% of papers
Papers with 10–49 citations	42	30% of papers
Papers with <10 citations	85	60.7% of papers
h-index of field	~15	Based on citation distribution

reflect moderate but growing maturity in the field, with ~5424 citations overall, a skewed average of ~39 citations per paper, a median of ~12, and an estimated h-index of ~15, revealing that most papers receive limited citations, while a small subset attracts significant attention.

Table 2. Impact by application area.

Application Area	Average Citations	Papers in Category	Key Technologies
Warehouse Robotics	181	5	Kiva, autonomous vehicles
Delivery Systems	158	4	Truck-drone coordination
Inventory Technology	45	52	RFID, computer vision, barcodes
Localization	32	18	UWB, LiDAR, SLAM
Path Planning	28	22	Optimization algorithms
Security/Surveillance	12	15	Computer vision, monitoring
Mapping/Modeling	8	10	3D reconstruction, digital twins
Multi-Robot Systems	15	14	UAV-UGV collaboration

presents impact differences across application areas, showing warehouse robotics, delivery systems, and inventory technologies as the most influential themes, leveraging Kiva-style autonomous vehicles, drone–truck coordination, RFID, and computer vision, while localization, path planning, multi-robot systems, security, and mapping/modeling contribute steadily but remain younger research directions within UAV-enabled warehouse automation.

3. Drone Technology in the Warehouse

Drones have become an inseparable part of warehouse operations due to the fact that they resolve many acute issues and offer powerful solutions. Drones increase operational efficiency by automating operations such as inventory management, stocktaking, and picking, which reduces the level of manual labor required and streamlines workflows. Their ability to gain access to small or difficult-to-access spaces improves inventory visibility and space utilization. The fact that drones also enable faster and more precise fulfillment of orders reduces the lead time and increases customer satisfaction. By providing the opportunity to respond proactively to the changing patterns of demand in real time by collecting and analyzing data, they facilitate decision-making. All in all, drones are changing the way of operating in warehouses by optimizing resource utilization, enhancing productivity, and providing businesses with a competitive advantage in the busy and challenging environment of contemporary supply chain management. The software and hardware components that combine to create drone technology in warehouse operations provide unmanned aerial vehicles (UAVs) with the ability to carry out a range of activities in a warehouse setting. These drones are able to execute some tasks remotely or automatically due to their advanced sensors, cameras, navigation, and communication protocols. Figure 6 lists the key aspects of drone technology that are utilized in warehouse operations.

3.1. Hardware Components

Several sophisticated hardware systems are installed on the drones to assist in the autonomous, efficient management of warehouses. Positioning and navigation sensors are required to navigate the drone around the warehouse. An example of this is LiDAR (Light Detection and Ranging), which aids in the extremely accurate mapping of the environment and also in the detection of obstacles. This ensures that it passes safely through busy, moving warehouses. LiDAR (Light Detection and Ranging) technology supports the mapping of surroundings and the detection of obstacles in a highly accurate manner. This ensures that it passes safely through busy, moving warehouses. Likewise, GPS provides the global positioning of the drone to track its location, and the IMU (Inertial Measurement Unit) detects the movement of the drone and its orientation so that it can be flown steadily through the problematic conditions in indoor locations. Ultrasonic sensors are generally utilized for proximity sensing of objects, and they help the drone avoid hitting surrounding objects. Drones have altimeters and barometers to maintain stability in the air and accurate height measurements, especially when maneuvering between various rack levels. Imaging systems allow the drone to interact with the inventory used in the warehouse. A vision camera is important for scanning barcodes and QR codes on inventory racks to guarantee efficient and accurate data capturing. Specific applications can also use a thermal camera, such as detecting hot spots in a warehouse or a machine that is malfunctioning by detecting its heat patterns. These imaging systems enable the drone to undertake some key inventory management functions precisely. The brain of the drone is an Autoflight controller, which takes care of the maneuvering as well as the stability of the flight and reaction to external influences. This is supplemented with the payload system, which provides the drone with object-handling capabilities such as a robotic gripper and an RFID reader to read the RFID tags on items in the inventory. Therefore, due to this hardware integration, drones can move lightweight objects in the warehouse and find inventory, increasing their operational efficiency.

3.2. Software and Communication

Robust software and communication infrastructure are significant in the success of drone operations in warehouse management. A Ground Control Unit (GCU) manages, monitors, and observes the activities of drones centrally. By connecting to the drone using specialized drone control software, users are able to program flight paths, monitor the drone in real time, and perform specific tasks such as scanning or moving objects. This software interface makes it possible to automate repetitive tasks and remain in control of operations. A good communication system makes it possible to exchange data between the drone and the ground control unit. This allows the drone to be fed instructions in real time and send live feedback, e.g., inventory and environmental information, to the control system. This real-time communication is necessary to maintain the operational accuracy of the work and react to any unexpected changes in the environment in the warehouse.

3.3. Working of Warehouse Management Drone

Through the integration of state-of-the-art hardware and software, warehouse management drones facilitate warehouse operations and the tracking of inventory. The initial step in the workflow of these drones is navigation and positioning [8]. Sensored drones, such as those using LiDAR and IMU (Inertial Measurement Unit), can autonomously navigate a warehouse. LiDAR allows the development of a three-dimensional map of the warehouse, which helps the drone recognize and avoid obstacles such as racks and people. At the same time, the IMU is able to check the orientation and movement of the drone to ensure that it stays stable during its flight, and GPS provides global positioning. These features enable the drone to fly safely in both stationary and dynamic warehouse settings. The drone communicates with the inventory through the use of imaging and scanning systems after it is airborne. The identification and cataloging of items in real-time is achieved with the help of a vision camera, which scans the barcodes and QR codes on the inventory racks. The RFID reader on the drone transmits the information to the warehouse database as soon as it scans the RFID tags on the inventory. This is highly enhanced with respect to the speed and accuracy of inventory tracking. Drones equipped with thermal cameras may also perform certain operations, such as the identification of heat spots within the warehouse to identify potential equipment failures or hazards.

Drone operations are supervised and controlled by a Ground Control Unit (GCU), which is a centralized point for controlling drones. Drone control software allows warehouse managers to delegate scanning, set specific flight paths, and monitor the drone in real time. This software reduces unnecessary flight time and saves power by making sure that the drone moves along the most efficient routes within the warehouse. An effective communication system provides real-time updates and a rapid response to any changes in the operations of the drone and the GCU to ensure that the data flow between the two is always active. Warehouse management drones may also deal with materials besides scanning inventory and inventory surveillance. Using a drone with a payload system, it is possible to transport small items, such as tools or small packages, to various areas of the warehouse. This aspect accelerates the flow of products in the plant and reduces the amount of manual labor. The drones may also be utilized in restocking shelves or delivering goods to required locations to enhance the efficiency of the operations. Lastly, the warehouse management system is connected to the data that is captured by the drone. When scanning QR codes, barcodes, and RFID tags, the drone ensures that the inventory is updated, there are fewer mistakes, and it enhances the accuracy of the stock levels. The presence of any differences, such as missing products or vacant shelves, can be recognized by visual inspection conducted by the drone using its cameras, and prompt corrective action can be taken by the managers. Figure 7 presents the working of the warehouse drone.

4. Applications of Warehouse Drones in Industrial Environments

There are many uses for warehouse drones that can greatly enhance productivity, precision, and general operations in storage facilities and distribution hubs. Among the main uses for drones in warehouses are the following [9].

4.1. Inventory Management (IM)

The use of self-propelled aerial vehicles loaded with advanced technologies to easily deliver control, surveillance, and management of stocks is referred to as drone use in stock management in warehouses. Traditional inventory management normally involves the application of labor-intensive and inaccurate fixed assets, such as forklifts and immobile scanners, in addition to manpower. On the other hand, drones operate with cameras, sensors, and software that allow them to perform tasks automatically (tracking, product recognition, and counting stocks). Drones are capable of traveling down aisles and even accessing hard-to-reach stacks in a warehouse, allowing them to work quickly and in real time and gather stock information that can instantly update the warehouse management system (WMS). This reduces costs, enhances the efficiency of operations, and reduces mistakes. Inventory management drones are fitted with high-resolution cameras, barcode scanners, and RFID readers. These drones travel along the aisle of the warehouse automatically using computer vision, LiDAR, or indoor positioning technology. After being placed, the drones scan the inventory on the shelf and obtain crucial data, such as location, quantity, and product ID. Once processed and received by the WMS, the data obtained is verified for accuracy and merged with the inventory records that are already present. In order to maximize productivity, a few drones are used during off-peak times and do not create disruptions to normal warehouse operations. Sophisticated drones with robotic arms can also be used to pick or sort lightweight items, allowing the inventory process to be further streamlined. The data from the drones is crucial in optimizing a warehouse. Drones help warehouse managers make decisions on whether items are overstocked or understocked by offering accurate and real-time inventory. This ensures that goods are replenished in time and reduces storage costs by avoiding unnecessary holding of excess stock. Moreover, drones help to collect spatial information regarding the utilization of shelves that can be utilized by managers to identify crowded or underutilized zones and streamline warehouse design. To allow faster access, items with high demand should be relocated towards dispatch areas, whereas those that are not used frequently should be stored in places that are not easily accessible. These insights allow decisions to be made based on the data, which enhances the work of the whole operation, reduces the retrieval times, and raises the level of storage efficiency.

The first attempts at automated inventory were made by Anderson [8], who modeled an automated warehouse management system with the help of UAVs. His model demonstrated a reduction in inventory by 40% and a reduction in manual labor by 32%, indicating the capabilities of UAV-assisted systems using MATLAB r2022b and RRT+ algorithms. On this basis, Sawant et al. [10] developed a powerful inventory control plan (RICS) with sliding mode control to track the UAV trajectory, which was more stable and efficient compared with traditional PID and MPC. The Kiva system by Ong et al. [11] installed more than 500 autonomous mobile robots to ferry inventory pods, and this revolutionized work in warehouses, enhancing the flexibility and throughput of the operations. Concurrently, Ehrenberg et al. [12] demonstrated the accuracy of RFID-equipped mobile robots in library inventory management, reaching centimeter-level precision, and Miller et al. [13] combined LIDAR and RFID to improve the ability to track assets over extensive distances. UAVs and UGVs were then used in practice by Guerin et al. [14] to scan a warehouse cooperatively, and Bae et al. [15] to scan a yard of yards with RFID technology. Beul et al. [16,17] developed these concepts with ROS-based SLAM and predictive control, which enabled micro aerial vehicles to work in 3D space autonomously. Mtita et al. [18] applied UAVs to logistics in airports, increasing the safety of their operations. At the same time, Yearling [19] and Fernandez-Caramesa et al. [20] proposed blockchain-powered frameworks for secure and transparent tracking, and Han et al. [21] proposed iBeacon-based schemes for outdoor asset monitoring. The idea of visual automation was accelerated by Cho et al. [22] and Xu et al. [23], as they introduced barcode and video-scan recognition algorithms that enhanced warehouses’ data collection and real-time processing, which was a transition from the simulative to the real-life level in UAV-based intelligent warehouses. Figure 8 below illustrates the system overview of the external management system (WMS).

By 2019, UAV-based warehouse management had entered a mature phase of experimentation and system integration. De Falco et al. [24] demonstrated UAV inventory systems using Region-based Convolutional Neural Networks (R-CNN) for autonomous visual stock tracking, while Fernández-Caramés et al. [25] combined blockchain and RFID to create secure, traceable Industry 4.0 supply chains. Jhunjhunwala et al. [26] compared UAV path-planning algorithms for efficient scanning and retrieval, and Macoir et al. [27] developed an ultra-wideband (UWB) localization system achieving centimeter-level accuracy. Multi-UAV coordination became prominent with Choi et al. [28] and Barlow et al. [29], who optimized multi-drone trajectories for faster audits using deep learning-based image analysis. In 2020, Cristiani et al. [30] designed compact mini-drones, combining blockchain data validation and wireless charging for uninterrupted operation, while Kwon et al. [31] improved localization through Extended Kalman Filter-based fusion. Digital transformation accelerated with Chen et al. [32], who developed a 5G-connected digital twin for remote UAV control, and Kalinov et al. [33], who optimized barcode detection through CNN-based vision models. Zhong et al. [34] enhanced UAV pose estimation via ArUco markers and Kalman filtering for accurate material handling. Together, these studies shaped modern UAV-based warehouse ecosystems, making them intelligent, networked, and adaptable, and laying the groundwork for autonomous, data-driven logistics systems.

Between 2021 and 2023, UAV-based research became increasingly collaborative and intelligent. Rhiat et al. [35] and Manjrekar et al. [36] automated barcode counting through Wi-Fi-enabled drones, while Guinand et al. [37] synchronized UAV and UGV fleets using ultra-wideband localization for efficient task allocation. Karamitsos et al. [38] reviewed drones’ benefits in real-time inventory tracking, and Li et al. [39] demonstrated precise RFID-based localization. Martinez-Martin et al. [40,41] and Yoon et al. [42] applied OCR and computer vision for indoor inventorying, while Liu et al. [43] used evolutionary algorithms to optimize UAV task planning. Fontaine et al. [44] introduced an embedded deep-learning RFID model with 93% accuracy, and Khropot and Volkanova [45] adapted drones for land parcel auditing. Subsequent research in 2022 deepened system precision and safety. Benes et al. [46] optimized UAV flight height for RFID reading in large warehouses, while Stanko et al. [47] demonstrated autonomous UAV navigation in real warehouses, overcoming visual-odometry and obstacle challenges. Collaborative ground–air frameworks appeared in Ribeiro et al. [48], formulating dual-vehicle path-planning for UAV-UGV inspection under budget constraints. Radácsi et al. [49] and Gubán & Udvaros [50] proposed GPS-free autonomous navigation. Najy et al. [51] explored truck–drone hybrid routing, and Proia et al. [52] focused on safe human–drone interaction. Advances by Li et al. [53] and Ekici et al. [54] improved localization accuracy, while Sales et al. [55] demonstrated faster multi-robot inventory operations.

Recent research from 2023 to 2024 emphasized optimization, multi-agent systems, and smart perception. Maweni et al. [56] improved flight efficiency by 38.6% using optimized zigzag routes, while Masnavi et al. [57] proposed a cooperative UAV–UGV model with a visibility-aware navigation framework. Lorenzo et al. [58] enhanced image-based accuracy to 91%, and Yang et al. [59] achieved 95% QR recognition with CNN-based methods. Qiu et al. [60] integrated lidar, IMU, and ArUco mapping for GNSS-denied environments, and the structure of their proposed work is shown in Figure 9. Emerging systems such as Kumar et al.’s [61] nano-drone model, Tsakiridis et al.’s [62] LiDAR-based pallet quantifier, and Lin et al.’s [63] reinforcement learning navigation showcased intelligent adaptability. Afandi et al. [64] accelerated inventory time from 15 to 2 min, and hybrid systems from Moreira et al. [65] and De Guzman et al. [66] improved energy efficiency in multi-robot coordination. Dias et al. [67] presented MANTIS with visual–inertial odometry, and Pawale et al. [68] integrated AI-based real-time data monitoring. Chen and Huang [69] introduced panoramic stitching for QR recognition, achieving high detection rates and improved automation accuracy. Overall, UAV-based inventory systems have evolved into highly coordinated, vision-driven, and efficient technologies that redefine modern warehouse management through automation, precision, and intelligent control.

Advantages: There are many benefits to using drones for inventory management in warehouses. These are given as follows:

• Compression of audit cycle time via autonomous coverage planning: UAVs reduce stocktaking time, primarily by converting aisle traversal into an automated coverage path problem, enabling high-frequency cycle counting (shift/night audits) without proportional labor scaling.
• Improved reachability and reduced access overhead in high-bay storage: UAV mobility enables direct sensing of elevated rack levels without forklifts/scissor lifts, reducing aisle disruption and exposure to fall hazards while preserving regular picking flow.
• Higher inventory data consistency through repeatable sensing pipelines: By standardizing flight speed, standoff distance, and scan geometry, drones produce more repeatable observations than manual scanning, lowering variance in counts and location confirmation.
• Near real-time discrepancy detection when connected to WMS: When barcode/QR/RFID observations are streamed to the WMS, mismatches (missing SKU, mis-slotting, wrong quantity) can be flagged as exceptions early, improving reconciliation latency and replenishment decisions.
• Scalable deployment with modular autonomy stacks: UAV inventory workflows can scale from small to large warehouses by adjusting mission segmentation and fleet size; the scaling driver is software (task allocation + path planning) rather than manpower.

Overall, drones transform the inventory process by bringing together speed, accuracy, and automation to enable the warehouse to keep up with today’s supply chain needs.

Limitations: Despite the drone’s advantages, several challenges or limitations exist in implementing drones for inventory management, which are given as follows:

• Endurance–coverage trade-off: Limited battery capacity constrains coverage per sortie, increasing dependency on docking, battery swap logistics, and mission partitioning; this becomes critical in large facilities and with high-frequency audit schedules.
• Identification reliability depends on modality and environment: Vision-based barcode/QR scanning is sensitive to glare, low illumination, occlusion, motion blur, and label orientation; RFID is sensitive to multipath and attenuation near metal racks. These effects can increase false negatives unless the sensing strategy is tuned.
• GPS-denied localization requirements: Dense indoor environments degrade GNSS, often requiring UWB/VIO/SLAM/marker-based localization; this adds infrastructure cost, calibration burden, and maintenance overhead.
• Collision and safety constraints in narrow aisles: Tight rack spacing plus dynamic obstacles (forklifts, workers) increases collision risk; safe deployment demands conservative speed planning and reliable obstacle sensing, which can reduce throughput.
• Integration overhead and workflow coupling: The benefit depends strongly on WMS integration (data schema, reconciliation logic, exception handling); without robust integration, drone counts may not translate into actionable inventory updates.

These challenges require careful planning and technological investment to ensure the effective use of drones in warehouse environments.

A detailed analysis of the applications of drone technology in inventory management has been presented in

Table 3. Detailed literature survey on the application of drone technology in inventory management.

Ref. No.	Year	Function	Robot Type	Single/Multiple Drones	Simulation Results	Real-Time Results	Technology/Methodology Used	Environment
11	2002	Inventory Management	UAV	Single	Yes	No	RRT*, MATLAB r2001b	Static
12	2005	Inventory Management	UAV	Single	Yes	No	RICS	Dynamic
13	2007	Inventory Management	UAV	Multiple	Yes	No	RFID	Static
14	2007	Inventory Management	Wheeled Robot	Single	Yes	No	AI and multi-robot coordination	Dynamic
15	2007	Inventory Management	Wheeled Robot	Multiple	Yes	No	RFID	Static
16	2009	Inventory Management	Wheeled Robot	Single	Yes	Yes	RFID, LiDAR, Rao-Blackwellized particle filter	Static
17	2016	Inventory Management	UAV + UGV	Single	Yes	No	RFID	Static
18	2016	Inventory Management	UAV	Single	Yes	Yes	RFID	Static
19	2017	Inventory Management	UAV	Single	Yes	Yes	ROS-based mapping and navigation	Static
20	2017	Inventory Management	UAV	Single	No	No	RFID	Static
21	2018	Inventory Management	UAV	Single	Yes	Yes	6D LiDAR-based localization	Static
22	2018	Inventory Management	UAV	Single	Yes	No	Autonomous vehicles, drones, digital inventory	Static
23	2018	Inventory Management	UAV	Single	No	Yes	Blockchain-based system	Static
24	2018	Inventory Management	UAV	Single	Yes	No	iBeacon-based ID reader	Static
25	2018	Inventory Management	UAV	Single	No	Yes	LBP, HOG, SVM classifier	Static
26	2018	Inventory Management	UAV	Single	Yes	No	Harris corner detector, Hough transform	Static
27	2019	Inventory Management	UAV	Single	Yes	No	R-CNN	Static
28	2019	Inventory Management	UAV	Single	Yes	No	Blockchain-based system	Static
29	2019	Inventory Management	UAV	Single	Yes	No	RFID	Static
30	2019	Inventory Management	UAV	Single	No	Yes	UWB localization system	Static
31	2019	Inventory Management	UAV	Multiple	Yes	Yes	Deep learning	Static
32	2019	Inventory Management	UAV	Multiple	Yes	No	Multi-UAV inventory tracking	Static
33	2020	Inventory Management	UAV	Multiple	Yes	No	Generic architecture	Static
34	2020	Inventory Management	UAV	Single	No	Yes	EKF-based multi-sensor fusion	Static
35	2020	Inventory Management	UAV	Single	–	–	Digital twin, 5G, cloud control	–
36	2020	Inventory Management	UAV	Single	Yes	Yes	CNN	Static
37	2020	Inventory Management	UAV	Single	Yes	No	Kalman filter-based pose estimation	Static
38	2021	Inventory Management	UAV	Single	–	–	ROS robots, RFID, bin-packing optimization	Static
39	2021	Inventory Management	UAV	Single	Yes	Yes	Barcode scanning with markers	Static
40	2021	Inventory Management	UAV + UGV	Multiple	Yes	No	UAV–UGV multi-robot system	Static
41	2021	Inventory Management	UAV	Single	Yes	No	Autonomous UAV routing, web interface	Static
42	2021	Inventory Management	UAV	Single	No	Yes	RFID relative localization	Static
43	2021	Inventory Management	UAV	Single	Yes	Yes	Computer vision techniques	Static
44	2021	Inventory Management	UAV	Single	Yes	No	QR segmentation, 3D localization	Static
45	2021	Inventory Management	UAV	Single	–	–	RFID drones, hybrid DE-Lion optimization	Static
46	2021	Inventory Management	UAV	Single	Yes	No	Hybrid differential evolution + RFID	Static
47	2021	Inventory Management	UAV	Single	Yes	No	Deep learning + RFID localization	Static
48	2021	Inventory Management	UAV	Single	Yes	Yes	Machine vision camera	Static
49	2022	Inventory Management	UAV	Single	–	–	RFID, RSSI-based altitude estimation	Static
50	2022	Inventory Management	UAV	Single	Yes	No	UHF RFID	Static
51	2022	Inventory Management	UAV	Single	Yes	No	Multi-robot systems	Static
52	2022	Inventory Management	UAV	Single	Yes	No	QR code	Static
53	2022	Inventory Management	UAV	Multiple	Yes	No	Genetic algorithm	Static
54	2022	Inventory Management	UAV	Single	Yes	No	MILP formulation	Static
55	2022	Inventory Management	UAV	Single	Yes	No	APF, LQR, iLQR	Static
56	2022	Inventory Management	UAV	Single	No	Yes	RFID + TCN	Static
57	2022	Inventory Management	UAV	Single	Yes	No	Virtual fiducial markers	Static
58	2023	Inventory Management	UAV	Multiple	Yes	No	Micro-drone + ground mobile robot	Static
59	2023	Inventory Management	UAV	Single	Yes	No	VACNA, CEM, CVAE	Static
60	2023	Inventory Management	UAV	Single	Yes	No	Visibility-aware cooperative navigation	Static
61	2023	Inventory Management	UAV	Single	Yes	No	Image capture and error classification	Static
62	2023	Inventory Management	UAV	Single	Yes	Yes	CNN-based QR reading	Static
63	2023	Inventory Management	UAV	Single	Yes	No	Visual mapping and volume estimation	Static
64	2024	Inventory Management	UAV	Multiple	Yes	Yes	Lightweight QR + dead reckoning	Dynamic
65	2024	Inventory Management	UAV	Single	Yes	No	LiDAR-driven approach	Static
66	2024	Inventory Management	UAV	Single	Yes	Yes	Reinforcement learning	Static
67	2024	Inventory Management	UAV	Single	Yes	Yes	Computer vision	Static
68	2024	Inventory Management	UAV	Multiple	Yes	Yes	VIO, SLAM, UWB, AprilTag	Dynamic
69	2024	Inventory Management	UAV	Multiple	Yes	Yes	SLAM and RFID	Static
70	2024	Inventory Management	UAV	Single	Yes	No	Visual odometry	Static
71	2024	Inventory Management	UAV	Single	No	Yes	Raspberry Pi and network integration	Static

, addressing parameters such as the reference number, year of publication, use of robots, use of single or multiple drones, testing in a simulation or a real-time environment, technology used, and environment (static and dynamic).

The analysis shown in Figure 10, Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16, Figure 17, Figure 18 and Figure 19 indicates that IM research growth is steady, with a sharp increase after 2018 and a peak during 2022–2024, showing an increase in interest in automated and intelligent warehouse systems (Figure 10 and Figure 16). Static environments are used in many works, while the dynamic warehouse scenario is targeted by only a few works with recent, gradual growth, pointing to a significant gap between experimental studies and those performed under real-world conditions (Figure 12;

Table 4. Research publication available on static and dynamic environment Year-wise distribution of research available on static and dynamic environments in IM.

Year Range	Static Environment	Dynamic Environment	Total
2002–2015	6	0	6
2016–2018	11	1	12
2019–2021	16	2	18
2022–2024	23	3	26
Total	56	6	62

). UAV-based platforms dominate research on IM; within that, single-UAV systems are used more frequently than multi-UAV, wheeled robots, and hybrid platforms, pointing to scalability as an open research challenge (Figure 11, Figure 13 and Figure 15;

Table 5. Research publication related to robot configuration Robot–environment configuration analysis.

Robot Type	Static Environment	Dynamic Environment	Total
Single UAV	42	4	46
Multiple UAVs	8	2	10
Wheeled Robot	4	0	4
UAV + UGV Hybrid	2	0	2
Total	56	6	62

and

Table 6. Research publication related to advanced technology adoption Technology adoption in different environments.

Technology	Static Environment	Dynamic Environment	Total
RFID	18	1	19
Computer Vision	15	2	17
	8	0	8
UWB	6	1	7
Blockchain	3	0	3
Other Technologies	6	2	8
Total	56	6	62

). Simulation-based validation is dominant, whereas real-time and hybrid validation approaches are comparatively fewer, especially in dynamic environments (Figure 14 and Figure 18;

Table 7. Research publication on real-world and simulation environment Real-world and virtual implementation analysis.

Environment	Simulation Only	Real-Time Results	Both (Simulation + Real-Time)	Total
Static	40	16	12	56
Dynamic	5	1	1	6
Total	45	17	13	62

). Technology trends point out the widespread use of RFID and computer vision, while emerging technologies include LiDAR, UWB, blockchain, and learning-based methods, which have been increasingly adopted in recent years with the aim of empowering autonomy, object-oriented accuracy, and data security (Figure 17 and Figure 19;

Table 8. Year-wise publication on different emerging technologies Chronological technology adoption.

Period	Dominant Technologies	Emerging Technologies
2002–2010	RFID, basic control algorithms	RRT*, RICS
2016–2018	LiDAR, ROS, blockchain	6D localization, digital twins
2019–2021	Deep learning, UWB, CNN	R-CNN, temporal networks
2022–2024	Reinforcement learning, VIO	VACNA, multi-agent systems

and

Table 9. Emerging technology and its function Functional capabilities by technology type.

Function	Primary Technologies	Accuracy Metrics	Real-Time Performance
Localization	UWB, LiDAR, VIO	5 cm–80 cm accuracy	Up to 2892 Hz update rate
Object Detection	CNN, R-CNN, SVM	87–95% detection accuracy	Real-time processing
Path Planning	RRT*, genetic algorithms	27% efficiency gain	Optimized trajectories
Data Processing	Blockchain, cloud computing	Secure and transparent	Real-time synchronization

). In general, the surveyed recent IM systems have good performance in localization, object detection, path planning, and real-time data synchronization, proving technical feasibility, while the focus should be on scalable, dynamic, and real-world validated solutions.

4.2. Maintenance and Inspections (MI)

Drones in warehouse maintenance and inspection utilize autonomous aerial vehicles for inspecting and evaluating the health of warehouse equipment, systems, and infrastructure for routine and on-demand inspection. These drones have on-board cameras that are of high resolution and use thermal sensor suites or other sophisticated equipment to detect issues in structures, machine failures, or other environmental conditions in real-time. Drones can navigate warehouse aisles with the help of cameras, thermal sensors, or LiDAR to capture a close-up view and temperature readings of high and narrow spaces. An example is that drones can be used to check roofs to identify any leaks, inspect the HVAC systems to ensure that they are functioning well, and also check the lights to determine whether they are functional. Some drones also have preset flight plans that guarantee even coverage of inspection sites, whereas others can use sophisticated indoor flight controls, such as SLAM (Simultaneous Localization and Mapping), to adapt their flight in dynamic environments in real time. The data obtained from the use of drones is analyzed using software or AI code that identifies any anomalies, such as cracks, overheating, or wear and tear, to facilitate proactive maintenance by the respective teams. Both high-resolution photographs and thermography are used to assist in diagnosing any possible problems in the structure of the warehouse, the electrical system, or overheating technology. Early detection of such problems prevents needless repairs, loss of production time, or occupational risks. Furthermore, drones generate large datasets in the long run, through which trends may be tracked, and maintenance needs may be predicted. As an example, a series of scans of a section of the roof can point to gradual deterioration and suggest an overhaul before much damage is inflicted. Data being used in maintenance ensures the maximization of resource management, planning, and cost control, thus maximizing overall efficiency in addition to the safety of the warehouse operations. Traditional methods of inspection and maintenance require manual work, heavy machinery, or ladder use, which may be time-consuming, physically strenuous, or dangerous. Such jobs are faster, safer, and more efficient with the help of drones, which can scan hard-to-reach locations such as upper shelves, roof construction, air-conditioning frameworks, or light assemblies.

The current advancement in UAV technologies has significantly increased their application in maintenance and inspection in various industries. Even at an earlier stage, projects like Itkin et al. [70] showed the strength of cloud-based monitoring systems, with real-time supervision of flights, collision avoidance, and multi-UAV control being managed by scalable server–client networks. These strategies indicated ways in which safe and reliable UAV operations in dynamic UAV environments could be facilitated through distributed computing, as indicated in Figure 20. Similarly, Eudes et al. [71], who conducted their research around the same time, demonstrated that autonomous mapping and navigation with micro aerial vehicles using only onboard sensors (stereo vision, inertial measurement units, and altimeters) was feasible, and no external infrastructure was required. Their combined efforts formed the foundation of an autonomous, real-time, and scalable UAV business in the industry. Based on these roots, the use of UAVs started to expand further into more narrow areas of application, in areas like manufacturing and the inspection of infrastructure. Mészáros et al. [72] suggested the use of drone-based systems to monitor processes, allowing the tracking of the moving objects with high precision under uncertainty, whereas Galar et al. [73] provided an extensive overview of autonomous vehicles when applied in the context of inspection processes and highlighted aspects such as their integration, practical usability, and safety in a real industrial environment. Their contributions to tools of inspection in warehouses, production lines, pipelines, and railways have shown that UAVs, ground robots, and hybrid systems can be used to complement one another. In the meantime, flight path optimization became one of the primary topics, and Tripicchio et al. [74] used model predictive control and SLAM-based localization to minimize mission time and deal with UAVs’ battery constraints. These techniques showed that algorithm development can contribute greatly to the efficiency and reliability of inspection.

Recent progress is directed at coordinated and large-scale activities with different UAVs. Batistatos et al. [75] proposed a swarm management platform that can coordinate UAV infrastructural fleets and area inspection, solving the issues of safety, communication, and resource allocation. Field experiments proved the viability of swarm-based design in dealing with large and complicated problems that can be covered by individual UAVs. When combined, these contributions represent an obvious evolution of individual UAV surveillance systems to fully autonomous micro-air vehicles, efficient flight path planning, and swarms of cooperative multi-UAVs. The trend underscores how UAVs are reshaping industrial inspection by offering greater autonomy, operational efficiency, and adaptability across both static and dynamic environments.

Advantages: The benefits of drones in maintenance and inspection are given as follows:

• Rapid access to elevated or constrained inspection zones: UAVs remove the need for scaffolding/lifts for roof, HVAC, lighting, and structural checks, enabling faster inspections with lower operational interruption.
• Higher inspection fidelity through multi-modal sensing: High-resolution imaging and thermal sensing enable early anomaly detection (overheating components, insulation loss, abnormal hotspots) and support preventive maintenance scheduling.
• Repeatable inspection trajectories for trend monitoring: Automated flight paths allow time-series comparison of the same asset region (same viewpoint and distance), making degradation detection more reliable than ad hoc manual inspections.
• Reduced downtime via off-peak autonomous routines: Scheduled autonomous runs can be executed during low-activity windows, minimizing disruption related to manual inspection procedures that often require area shutdown.
• Digitized evidence and traceability: UAV-collected imagery and thermal maps provide auditable records that support maintenance reporting, compliance, and root-cause analysis.

Limitations: Despite their benefits, drones have several challenges in warehouse maintenance and inspection, which are given as follows.

• Indoor navigation complexity in cluttered industrial environments: Close-proximity inspection requires stable position holding and precise obstacle avoidance near racks and machinery; this typically demands SLAM/VIO and robust control, raising system complexity.
• Sensor payload and cost escalation: Thermal cameras, high-quality optics, and stabilization hardware increase payload and platform cost and may shorten endurance, creating a cost–capability trade-off.
• Limited flight time constrains inspection completeness: Large warehouses may require multiple sorties; frequent docking/charging becomes a scheduling bottleneck.
• Data management burden: Continuous visual/thermal capture produces large datasets; storage, labeling, and analytics pipelines are non-trivial and can become the limiting factor rather than flight.
• Safety and privacy compliance: Indoor UAV operation must meet site safety policies (human proximity, rotor safety, noise) and privacy rules when cameras operate near personnel or sensitive operations.

These challenges need cautious planning and investment, as well as adaptation, in order for drones to unlock their full potential for maintenance as well as inspection.

The consolidated MI research analysis in Figure 21, Figure 22, Figure 23, Figure 24, Figure 25, Figure 26, Figure 27, Figure 28, Figure 29 and Figure 30 and

Table 10. Detailed literature survey on the application of drone technology in maintenance and inspection.

Ref. No.	Year	Function of Drone	Robot Type	Single/Multiple Drones	Simulation Results	Real-Time Results	Technology/Methodology Used	Environment
72	2016	Maintenance and inspection	UAV	Multiple	Yes	No	Cloud-based web application	Dynamic
73	2017	Maintenance and inspection	UAV	Single	Yes	Yes	Onboard perception, localization, safe navigation	Static
74	2018	Maintenance and inspection	UAV	Single	Yes	No	Indoor localization, closed-loop tracking	Static
75	2020	Maintenance and inspection	UAV, UGV	Multiple	No	No	Autonomous vehicles, system integration, case studies	Static and dynamic
76	2023	Maintenance and inspection	UAV	Single	Yes	Yes	Model predictive control and SLAM	Dynamic
77	2024	Maintenance and inspection	UAV	Single	Yes	No	UAV swarm management technique	Dynamic

shows clear technological progress in robotic platforms, environments, and tools. Single-UAV systems dominate (Figure 21 and Figure 27), with limited use of multi-UAV or UAV–UGV hybrids. There is a noticeable shift from static to dynamic environments over time (Figure 22,

Table 11. Year-wise distribution of research available on static and dynamic environments in MI.

Year Range	Static Environment	Dynamic Environment	Total
2016–2017	1	1	2
2018–2020	1	1	2
2023–2024	0	2	2
Total	2	4	6

). Most studies still rely on simulations rather than real-world tests (Figure 23 and Figure 24. Research activity has gradually increased (Figure 25 and Figure 28), aligned with evolving technologies (Figure 26) and adoption trends (Figure 29 and Figure 30). Single UAVs are used in both static and dynamic settings, while advanced setups are tested mainly in dynamic scenarios (

Table 12. Robot–environment configuration analysis of MI.

Robot Type	Static Environment	Dynamic Environment	Total
Single UAV	2	2	4
Multiple UAVs	0	1	1
UAV + UGV Hybrid	0	1	1
Total	2	4	6

and

Table 13. Technology adoption in different environments in MI.

Technology	Static Environment	Dynamic Environment	Total
Cloud computing	0	1	1
Perception and navigation	1	1	2
Localization systems	1	0	1
Vehicle integration	0	1	1
Advanced control	0	1	1
Total	2	4	6

). A shift toward advanced control, SLAM, and swarm systems is evident (

Table 14. Real-world and virtual implementation analysis of MI.

Environment	Simulation Only	Real-Time Results	Both (Simulation + Real-Time)	Total
Static	2	1	0	3
Dynamic	2	1	0	3
Total	4	2	0	6

and

Table 15. Chronological technology adoption in MI.

Period	Dominant Technologies	Emerging Technologies
2016–2017	Cloud computing, basic perception systems	Web-based drone applications, onboard navigation
2018–2020	Localization systems, vehicle integration	Closed-loop tracking, multi-vehicle coordination
2023–2024	Advanced control (MPC), SLAM, swarm management	Model predictive control, swarm intelligence algorithms

), though real-time readiness is higher in perception and navigation than in multi-robot or cloud-based systems (

Table 16. Functional capabilities by technology type in MI.

Function	Primary Technologies	Accuracy Metrics	Real-Time Performance
Localization and positioning	Indoor localization, closed-loop tracking, IR markers	Centimeter-level precision, sub-meter accuracy	Medium (limited real-time validation)
Perception and navigation	Onboard perception, safe navigation, vision-based methods	Object detection accuracy, collision avoidance reliability	High (proven real-time capability)
Advanced control	Model predictive control (MPC), SLAM	Predictive accuracy, mapping precision	High (100% real-time success rate)
Multi-robot coordination	UAV swarm management, vehicle integration	Coordination efficiency, task allocation success	Low (simulation-only, no real-time results)
Cloud and distributed systems	Cloud-based web applications	Processing latency, data synchronization	Low (no real-time implementation)

). Overall, the data reflects rising technological maturity but highlights ongoing gaps in real-world multi-robot deployment and validation.

4.3. Picking and Packing or Goods Transportation (PPGT)

Drone-based picking and packing in warehouses is the use of drones for the purpose of finding, extracting, and moving items from storage racks to packing or shipment areas. Manual picking and packing in conventional warehouses involve workers physically moving through the storage rows, finding articles, and staging them for shipment. Drones simplify these steps through automation of the extraction and conveyance of materials, mainly lighter items, with less time and effort. Picking and packing drones are fitted with specialized grippers, such as robotic grippers, suction devices, or clamps, with which they can seize and transport items. These drones move within the warehouse via indoor positioning systems, LiDAR, or computer vision to identify the necessary stock. Once an item is located, it is picked from its storage position and transported to the packing or sorting station. The WMS functions as the controller throughout the process, sending real-time item position as well as order requirement data to the drone. Certain drones are also able to work in cooperation with human employees or floor-based robots, delivering items for packing purposes with decreased human effort. Drones can also work in non-peak hours with the goal of boosting productivity as well as reducing interference in routine warehouse activities. Drones collect data that gives warehouse managers useful insights that can maximize picking and packing activities. For example, drones capture real-time item locations and stock quantities, allowing for enhanced tracking of inventories and having frequently picked items available. The data allows managers to analyze order-fill patterns and detect bottlenecks, as well as optimize storage layouts for faster access to in-demand items.

The studies conducted over the last ten years have demonstrated the capabilities of UAVs to enhance logistics and materials management within warehouses and factories. Similarly to the cases of Olivares et al. [76] and Cordova and Olivares et al. [77], early simulation-based experiments investigated how fleets of drones can deliver semi-finished products and control the inventory processes of production facilities. Their models emphasized the impact of the UAV capacity, cycle times, and demand variability on the efficiency of the fleet, which implied that fleets could keep production flowing and minimize bottlenecks. Like Carlsson and Song et al. [78], Derpich et al. [79] used a combination of mathematical modeling and routing optimization to demonstrate that the use of UAVs with trucks or power-efficient path planning could reduce delivery times and save power in the transportation of goods. The main themes behind these early works were mostly simulation and theoretical modeling, but they provided important foundations for UAV-assisted material flow in manufacturing and logistics. An example of one such attempt to deploy UAV swarms in logistics was proposed by Kuru et al. as a smart platform [80]. Their paper looks into the ways in which UAV swarms can be effectively deployed to deliver goods through a comparison of various schemes and routing schemes in the context of various operation conditions. It presents a dynamic task-assignment methodology that assists in assigning drones and 3D routes in warehouses more efficiently. The results of simulations using different data sets indicate that the decision of the method of delivery is heavily dependent on the nature of the data and constraints of the system. This research could be useful in guiding aviation authorities, but it also conceptualizes the future regulatory framework for drone-based logistics. It also gives advice to the manufacturers of UAVs and logistics companies developing large-scale drone delivery systems.

As the interest in practical applications increased, subsequent research started experimenting with UAV-based systems in real-life or semi-controlled settings. The authors Lieret et al. [81], Sorbelli et al. [82], and Kloetzer et al. [83] designed an integrated drone and load-handling system and algorithm to optimize UAV picking and routing under the conditions of a warehouse that is not in motion. Their findings indicated that drones would be more effective than the traditional worker-based picking for some layouts and demand patterns, whereas mathematical programming may reduce the battery consumption and travel time in a complex facility. More recent research has focused on scalability, cost, and integration with smart warehouse operations. A low-cost autonomous robot introduced by Alam et al. [84] to be used by small and medium enterprises demonstrated that even resource-limited environments could be improved with the help of automation. Ham et al. [85] also showed that drone fleets can optimize large quantities of transfers in warehouses in real time, which indicates the possibility of throughputs in high-demand fulfillment centers. Taking the idea of futuristic applications, Jeong et al. [86] reviewed the proposed airborne fulfillment centers by Amazon, where optimization models are applied to plan the fleets of aerial warehouses and their UAVs. Han et al. [87] expanded on this with dynamic order-picking approaches, demonstrating that real-time path reallocation and multi-UAV coordination could greatly minimize travel distances and completion times. Together, these papers point to the direction of simulation-based feasibility research for scalable, AI-assisted, and real-time UAV tasks in industrial logistics and warehouse automation (Figure 31).

Advantages: Drones offer several benefits in picking and packing operations, which are given as follows:

• Reduced internal transport latency for light-load workflows: UAVs can shorten the time required to move small parts, tools, and documents between stations by bypassing ground congestion and reducing human walking time.
• Task automation through WMS-driven dispatching: When orders trigger UAV missions directly, retrieval/delivery can be coordinated with packing and sortation workflows, improving responsiveness during peak demand.
• Lower congestion around picking aisles (in suitable layouts): For selected facility designs, UAVs can reduce ground traffic and aisle interference by shifting micro-transport tasks into the air.
• Precision delivery to defined drop points: Controlled landing or hover-drop procedures can improve repeatability of “handoff” to packing stations (especially in tightly timed production logistics).
• Potential for multi-agent throughput scaling: In principle, fleet scheduling enables parallelism, where multiple UAVs handle simultaneous micro-deliveries to reduce queueing delays.

Limitations: Despite their advantages, drones face several challenges in picking and packing operations, given below.

• Payload and grasping constraints: Most UAV systems remain limited to lightweight items and require reliable gripping mechanisms; heavy/bulky or fragile goods still demand ground robots or humans.
• Safety constraints near humans and inventory: Airborne movement near workers introduces risk; safe speeds, separation distances, and protective hardware may reduce practical throughput.
• Fleet coordination remains under-validated in real warehouses: Many studies demonstrate routing/scheduling in simulation; real-world multi-UAV coordination faces communication delays, localization drift, and collision avoidance complexity.
• Endurance limits under frequent transport cycles: Repeated pick–carry–deliver loops create high energy consumption; frequent recharging or swap cycles must be engineered into operations.
• Reliability depends on perception and handling accuracy: Incorrect identification, grasp slippage, or unstable hover can cause mis-picks, product damage, or operational stoppage, requiring robust perception and control.

The paper provides a detailed review of drone-based picking, packing, and goods transportation (PPGT) systems in warehouse environments, showcasing findings in Figure 32, Figure 33, Figure 34, Figure 35 and Figure 36 and

Table 17. Detailed literature survey on the application of drone technology in picking and packing or goods transportation.

Ref. No.	Year	Function of Drone	Robot Type	Single/Multiple Drones	Simulation Results	Real-Time Results	Technology /Methodology Used	Environment
78	2015	Picking, packing, and goods transportation	UAV	Multiple	Yes	No	Simulation model, multi-drone logistics	Dynamic
79	2016	Picking, packing, and goods transportation	UAV	Multiple	Yes	No	Fleet management model for VTOL-UAVs	Dynamic
80	2018	Picking, packing, and goods transportation	UAV	Multiple	Yes	No	Hybrid delivery system using truck and UAV	Static
81	2018	Picking, packing, and goods transportation	UAV	Multiple	Yes	No	3D vehicle routing, heuristic optimization	Static
82	2019	Picking, packing, and goods transportation	UAV	Single	Yes	No	Dimensionality reduction, Hungarian and cross-entropy Monte Carlo methods	Static
83	2019	Picking, packing, and goods transportation	UAV	Single	Yes	No	Autonomous navigation, flight tracking, energy-efficient load handling	Static
84	2019	Picking, packing, and goods transportation	UAV	Single	Yes	Yes	Christofides algorithm	Static
85	2019	Picking, packing, and goods transportation	UAV	Single	Yes	Yes	Mathematical programming, routing, real-time optimization	Static
86	2020	Picking, packing, and goods transportation	UAV	Single	Yes	No	Mathematical programming-based planning and simulation	Static
87	2020	Picking, packing, and goods transportation	UAV	Single	Yes	No	Mixed integer programming and constraint programming	Static
88	2022	Picking, packing, and goods transportation	UAV	Multiple	Yes	No	Mixed integer linear programming (MILP) model	Dynamic
89	2022	Picking, packing, and goods transportation	UAV	Multiple	Yes	No	Ant colony optimization and k-opt-based algorithms	Dynamic

,

Table 18. Year-wise distribution of research on static and dynamic environments in PPGT.

Year Range	Static Environment	Dynamic Environment	Total
2015–2016	0	2	2
2017–2018	2	0	2
2019–2020	6	0	6
2021–2022	0	2	2
Overall Total	8	4	12

,

Table 19. Robot–environment configuration analysis in PPGT.

Robot Type	Static Environment	Dynamic Environment	Total
Single UAV	6	0	6
Multiple UAVs	2	4	6
Total	8	4	12

,

Table 20. Technology adoption in different environments in PPGT.

Technology/Method Used	Static Environment	Dynamic Environment	Total
Simulation-based logistics models	0	1	1
Fleet management (VTOL)	0	1	1
Hybrid delivery system (truck + UAV)	1	0	1
3D vehicle routing and heuristics	1	0	1
Dimension reduction, Hungarian, and CEMC	1	0	1
Autonomous navigation with lightweight load system	1	0	1
Christofides algorithm	1	0	1
Mathematical programming (routing and planning)	2	0	2
Mixed integer programming and constraint programming	1	0	1
MILP model	0	1	1
Ant colony optimization and k-opt	0	1	1
Total	8	4	12

,

Table 21. Real-world and virtual implementation analysis in PPGT.

Environment	Simulation Only	Real-Time Results	Both (Simulation + Real-Time)	Total
Static environment	6	0	2	8
Dynamic environment	4	0	0	4
Overall total	10	0	2	12

,

Table 22. Chronological technology adoption in PPGT.

Period	Dominant Technologies	Emerging Technologies
2015–2016	Simulation-based logistics; VTOL fleet management	Multi-UAV coordination models
2017–2018	3D routing; heuristic optimization	Hybrid delivery systems
2019	Mathematical programming; Christofides algorithm; autonomous navigation	Cross-entropy Monte Carlo; advanced dimensionality reduction
2020	Mixed integer programming; constraint programming	Real-time routing integration
2021–2022	MILP models; ant colony optimization (ACO)	k-opt optimization; swarm-oriented MILP

and

Table 23. Functional capabilities by technology type in PPGT.

Function	Primary Technologies	Accuracy Metrics	Real-Time Performance
Routing and path planning	Heuristic algorithms, Christofides algorithm, MILP, ACO	Route optimality, path length deviation	Limited real-time validation in selected 2019 studies
Navigation and tracking	Autonomous navigation systems, onboard perception	Localization accuracy, tracking precision	Partially real-time (reported mainly in 2019)
Load handling and delivery	Lightweight UAV load devices, hybrid truck–UAV systems	Load balance, payload efficiency	Mostly simulation-based
Fleet and swarm coordination	VTOL fleet models, multi-drone logistics, MILP	Scheduling efficiency, conflict avoidance	No real-time demonstrations
Optimization and planning	Integer programming, constraint programming, k-opt	Convergence speed, optimal cost	Simulation-only

. A clear trend emerges from the early reliance on static environments toward recent experimentation in dynamic scenarios (Figure 32; Table 18). although the majority of implementations still occur in simulated settings rather than real-world applications (Figure 35; Table 21). Single UAVs dominate static deployments, while multi-UAV systems are more common in dynamic settings (Figure 33; Table 19), reflecting a gradual shift toward more complex coordination strategies. The timeline of technological adoption (Figure 36; Table 20 and Table 22) reveals the early dominance of simulation-based logistics and routing algorithms, followed by the integration of advanced methods like MILP, ACO, and swarm intelligence in later years. Despite these advancements, real-time results remain scarce, with only limited demonstrations of practical deployment across both static and dynamic contexts. Functional analysis (Table 23) shows strong simulation performance in routing, navigation, and planning, but little real-time validation for fleet coordination and load delivery. Overall, while research activity is increasing (Figure 34), the field still faces major gaps in real-time readiness and scalability, underscoring the need for further development and deployment of multi-UAV systems in live warehouse environments.

4.4. Security and Surveillance (SS)

Security and surveillance of warehouses through the use of drones involves the operation of autonomous or remotely piloted aerial vehicles for the purpose of watching over and protecting the warehouses’ premises. The drones are equipped with high-definition cameras, thermal imaging, and motion detection, as well as other surveillance devices for the detection of unauthorized persons, tracking assets, and ensuring adherence to safety measures. Conventional security systems depend on fixed cameras, human security teams, or alarm systems, which are subject to blind spots or limited coverage areas. Drones upgrade security operations through real-time aerial coverage and quick patrolling of expansive areas, as well as improving the response to security alert incidents. The drones patrol warehouse premises, inside or out, along a set route, or react to a perceived danger. Drones can operate in the dark and in low light conditions with night vision cameras and by detecting thermal changes, thus identifying intruders or suspicious movements. To implement these systems in the interior, the drones go over aisles and patrol limited or valuable storage spaces to enforce access processes. On the outside, drones scan the borders, address doorways and exits, and search for violations or questionable activity. Drone technology has been further developed to be combined with alarm systems and WMS, providing automatic reactions, e.g., alerting guards or closing doors. Emergencies, including fire or unauthorized access, can also be communicated by using drones to immediately check the situation and provide live video streams to the responding agencies to improve the efficiency of their actions. Sensors on drones are able to inspect the temperature and humidity of warehouses, among other environmental factors. This is very important in industries that require controlled storage conditions, such as for pharmaceuticals and perishable products.

The evolution of UAVs in the last ten years has been consistent, from single-robot systems that are used in tracking inventory and security to more sophisticated systems that can perform complicated industrial procedures. The initial examples, including the work of Jyothsnaa et al. [89] and Lioulemes et al. [88], covered wheeled robots and UAVs with image processing and AI used to automate the process of counting inventory in warehouses, cargo space optimization, and the safe navigation of robots around people in indoor environments. Subsequently, attempts at making drone activity more precise and coordinated, such as in Kalinov et al. [90], where the UAVs could successfully localize themselves and land on mobile robots, and Dawaliby et al. [91], who implemented blockchain-based systems to manage drones in the IoT, were made. These works demonstrated the potential of drones to be deployed autonomously and in a structured setting, which provided the precursors of safe, smart, and dependable inventory and inspection solutions. An example of this is the work by Anwar et al. [92]. His paper introduces a fully automated system that utilizes drones and UAVs to conduct real-time construction monitoring and reporting through the creation of 3D photogrammetric models and compares them with BIM data. This system greatly saves manpower and improves the tracking of progress, verification of bills, and accuracy of plans at the site.

Since 2021, the development of research has shifted to the inclusion of UAVs in more dynamic and multi-purpose applications. Reports like Oommen et al. [93] and Alsayed et al. [94] displayed hybrid UAV-rovers performing disaster rescue, or LiDAR being used on an industrial stockpile using a drone (Figure 37). Yi and Qu [95] mentioned the potential of drones in the monitoring of construction sites, and Stanko et al. [96] and Martinez-Carranza and Rojas-Perez [97] demonstrated the possibility of using UAVs to monitor warehouses with a minimal number of operators by employing SLAM, QR detection, and human–machine interfaces. Since then, Awasthi et al. [98] and Kamal et al. [99] applied these ideas to multi-UAV and UAV-UGV multi-agent scenarios and found that drones were able to deliver packages safely to humans and facilitate coordinated activities during surveillance operations. Such publications focused on the flexibility of UAVs, incorporating perception, localization, and real-time communication to increase their utility in Industry 4.0 systems.

The latest literature provides a solid indication of a transition to large-scale, AI-driven, and edge-computing-enabled drone systems. Multi-UAV swarms with blockchain and deep learning to control inventory and security in smart factories were proposed by Masuduzzaman et al. [100], whereas IoT-enabled UAVs to inspect infrastructures to improve proactive maintenance were demonstrated by Kumar et al. [101]. Mourya et al. [102] applied swarm behavior to disaster monitoring. Other refinements of UAVs include those of Mezouari et al. [103], who introduced low-power neural accelerators into their devices to assist in real-time surveillance activities that do not significantly impact flight time and detection rates. All of that proves that single-UAV prototypes have quickly evolved into multi-drone and AI-powered systems that include IoT and are becoming increasingly involved with the automation of warehouses, industrial inspection, construction monitoring, and disaster management.

Advantages: Drones have several benefits for surveillance and security in warehouses, given as follows:

• Adaptive coverage beyond fixed-camera blind spots: UAVs provide mobile viewpoints and can re-route patrol paths based on alerts, addressing occluded zones that fixed CCTV cannot cover reliably.
• Faster incident verification through real-time aerial feed: During alarms (intrusion, fire risk), UAVs can deliver immediate visual/thermal confirmation, improving response prioritization and decision-making.
• Enhanced sensing capability with thermal/night vision: In low-light or smoky conditions, thermal imaging can detect anomalies that visible cameras miss, supporting early hazard detection.
• Reduced dependence on continuous human patrol: Semi-autonomous patrol missions reduce repetitive monitoring workload and allow guards to focus on interventions rather than constant inspection.
• Environmental condition monitoring: UAV-mounted sensors can monitor temperature/humidity patterns in sensitive storage areas, enabling compliance monitoring and early detection of abnormal conditions.

Limitations: Drones face several challenges in security as well as surveillance, which are mentioned below.

• False alarms and perception uncertainty: Vision-based detection is sensitive to shadows, reflections, moving machinery, and occlusions, which can increase false positives without robust models and calibration.
• Privacy and data governance constraints: Continuous imaging can capture sensitive operational details; strict policies and data handling protocols are required to avoid compliance violations.
• Localization and obstacle avoidance in dynamic indoor spaces: Reliable indoor navigation is difficult in the presence of moving forklifts and workers; safe operation increases autonomy complexity and cost.
• Endurance limits for continuous patrol: Persistent surveillance requires fleet rotation, charging strategy, and uptime planning; otherwise, coverage continuity suffers.
• Cybersecurity risks: UAV communication links and onboard compute can become attack surfaces; secure communication and access control are needed for industrial deployment.

The field of security and surveillance (SS) has seen a notable shift in technological approaches and research focus over the years. Figure 32, Figure 33, Figure 34, Figure 35, Figure 36, Figure 37, Figure 38, Figure 39, Figure 40, Figure 41, Figure 42, Figure 43, Figure 44, Figure 45 and Figure 46 and

Table 24. Detailed literature survey on the application of drone technology in security and surveillance.

Ref. No.	Year	Function of Drone	Robot Type	Single/Multiple Drones	Simulation Results	Real-Time Results	Technology/Methodology Used	Environment
90	2010	Security and surveillance	Wheeled robot	Single	Yes	No	AI algorithms, image processing, path planning	Static
91	2014	Security and surveillance	UAV	Single	Yes	No	Vision-based navigation, pedestrian detection	Static
92	2019	Security and surveillance	UAV	Single	Yes	No	Sensor fusion, IR markers, localization	Static
93	2020	Security and surveillance	UAV	Single	Yes	Yes	Blockchain platform, Z-score anomaly detection	Static
94	2018	Security and surveillance	UAV	Single	Yes	Yes	Photogrammetry techniques	Dynamic
95	2021	Security and surveillance	UAV	Single	Yes	Yes	Hybrid UAV with integrated rover	Dynamic
96	2021	Security and surveillance	UAV	Single	Yes	Yes	LiDAR sensors and 3D static scanners	Dynamic
97	2021	Security and surveillance	UAV	Single	No	No	Drone-based visual data acquisition and dynamic planning	Dynamic
98	2022	Security and surveillance	UAV	Single	Yes	No	Autonomous drone with human–machine interface	Dynamic
99	2022	Security and surveillance	UAV	Multiple	No	Yes	UAV swarm, motion capture system, ROS, Unity	Dynamic
100	2022	Security and surveillance	UAV	Single	Yes	No	Metric monocular SLAM	Static
101	2023	Security and surveillance	UAV	Multiple	Yes	Yes	UAV-assisted multi-robot system	Dynamic
102	2024	Security and surveillance	UAV	Multiple	Yes	No	Multi-UAV system	Dynamic
103	2024	Security and surveillance	UAV	Single	Yes	Yes	Drone technology and IoT integration	Dynamic
104	2024	Security and surveillance	UAV	Multiple	Yes	No	Multiple autonomous drones	Dynamic
105	2024	Security and surveillance	UAV	Single	Yes	Yes	Raspberry Pi 4B with Intel Neural Compute Stick 2 VPU	Dynamic

,

Table 25. Year-wise distribution of research on static and dynamic environments.

Year	Static Environment	Dynamic Environment	Total
2010	1	0	1
2014	1	0	1
2018	0	1	1
2019	1	0	1
2020	1	0	1
2021	0	3	3
2022	1	2	3
2023	0	1	1
2024	0	4	4
Total	5	11	16

,

Table 26. Robot–environment configuration analysis.

Robot Type	Static Environment	Dynamic Environment	Total
UAV	4	11	15
Wheeled robot	1	0	1
Total	5	11	16

,

Table 27. Technology adoption in different environments.

Technology/Method Used	Static Environment	Dynamic Environment	Total
AI algorithms, image processing, path planning	1	0	1
Vision-based navigation, pedestrian detection	1	0	1
Sensor fusion, IR markers, localization	1	0	1
Blockchain platform, Z-score anomaly detection	1	0	1
Photogrammetry techniques	0	1	1
Hybrid UAV with integrated rover	0	1	1
LiDAR sensors and 3D static scanners	0	1	1
Drone-based visual monitoring and dynamic planning	0	1	1
Autonomous drone with human–machine interface	0	1	1
UAV swarm with motion capture, ROS, and Unity	0	1	1
Metric monocular SLAM	1	0	1
UAV-assisted multi-robot system	0	1	1
Multi-UAV system	0	1	1
Drone and IoT integration	0	1	1
Raspberry Pi 4B with Intel Neural Compute Stick 2 VPU	0	1	1
Total	5	11	16

,

Table 28. Real-world and virtual implementation analysis.

Environment Type	Simulation Only	Real-Time Only	Both (Simulation + Real-Time)	Total
Static	4	0	1	5
Dynamic	4	2	5	11
Total	8	2	6	16

,

Table 29. Chronological technology adoption.

Period	Dominant Technologies	Emerging Technologies
2010–2014	AI-based surveillance, vision navigation, IR markers	Early SLAM, lightweight path planning
2015–2019	Indoor localization, photogrammetry, cloud tools	Blockchain, hybrid UAV–rover systems, LiDAR scanning
2020–2022	Multi-UAV systems, motion capture with ROS, HMI-based UAVs	IoT-enabled drones, advanced SLAM, Unity-based virtual testing
2023–2024	UAV-assisted multi-robot systems, dynamic planning	Edge AI (Intel NCS2), swarm intelligence

and

Table 30. Functional capabilities by technology type.

Function	Primary Technologies	Accuracy Metrics	Real-Time Performance
Surveillance and security	Vision processing, AI-based detection, path planning	Object detection accuracy, false alarm rate	Low-latency video streaming and stable tracking
Navigation and localization	SLAM, IR markers, LiDAR, sensor fusion	Pose estimation error, drift rate	Real-time mapping and localization
Dynamic planning and control	Blockchain validation, Z-score anomaly detection, predictive control	Anomaly detection success rate	High responsiveness under dynamic conditions
Multi-robot coordination	Swarm algorithms, ROS with motion capture, Unity simulation	Robot synchronization accuracy	Real-time fleet coordination
Human–machine interaction	HMI-based UAV, autonomous control	Response time, user command accuracy	Direct real-time control execution
Edge AI processing	Raspberry Pi 4B with Intel NCS2, onboard inference	Inference accuracy, FPS performance	High-speed onboard decision-making

collectively show a growing preference for dynamic environments over static ones, with 11 out of 16 recent studies targeting dynamic scenarios (Table 25; Figure 39), and UAVs dominating as the primary robotic platform (Figure 38; Table 26). Research activity in SS has steadily increased (Figure 40), especially post-2020, aligning with a surge in multi-UAV systems (Figure 41) and edge AI adoption (Table 27, Table 29 and Table 30; Figure 43, Figure 44 and Figure 45). Real-time implementation remains limited, as simulation environments still dominate, particularly in dynamic contexts (Table 28; Figure 42). Across studies, advanced technologies such as SLAM, ROS, Unity-based simulations, and AI-driven path planning are pivotal for real-time performance, accurate detection, and multi-robot coordination (Table 30; Figure 46). These trends indicate a clear movement toward intelligent, adaptive, and collaborative robotic systems in increasingly complex environments.

4.5. Warehouse Layout Optimization (WLO)

Drones play an important role in warehouse layout optimization, as they capture insights and data that optimize the spatial distribution of stock, equipment, and routes in a warehouse. An optimized layout reduces unnecessary movement, eliminates inefficiencies in operations, and optimizes worker safety. Conventional layout optimization is performed through manual surveys as well as static data, both of which can be time-consuming and inaccurate. Drones change the equation through capturing detailed images, 3D blueprints, and live data, providing a dynamic and accurate view of operations in the warehouse in order to aid in layout amendments as well as enhancements. Drones with cameras, LiDAR scanners, and mapping capabilities are dispatched to scan the warehousing environment, constructing precise spatial models of the layout. The drones move through the corridors, around the shelves, and over the storage areas, capturing data about inventory locations, workflow congestion points, and vacant areas. The captured data is analyzed to create 3D models or digital twins of the warehousing facility, which helps the warehousing managers maintain an interactive, accurate visual of the facility. The drones can also monitor worker movement, equipment, and goods patterns, enabling managers to observe how the existing layout contributes to productivity. With these insights, managers can restructure shelves, identify ideal paths, and distribute storage areas in an efficient way to improve operations.

Inventory control and industrial applications of UAVs have continued to increase in sophistication from early demonstration systems to highly complex systems. Preliminary projects like Parkison et al. [104] and Olivares et al. [105] showed the potential of drones to be integrated with RFID and optimization models to be used in warehouse mapping, route planning, and internal logistics. Fan et al. [106] also demonstrated that low-priced quadcopters equipped with vision-based mapping were capable of centimeter-level accuracy when mapping a warehouse layout. In the development of the field, scholars started dealing with the shortcomings of centralized control and scalability. Distributed task allocation was presented by Kattepur et al. [107] as a means of multi-robot control and by Elmokadem [108] as a means of precision, provided that decentralized control could facilitate flexibility, collision avoidance, and real-time adaptability. Taken together, these articles underscore the increased functionality of drones in the field of logistics and inventory management, especially in an orderly agriculturesetting.

By 2019–2021, UAV technology usage had reached various industrial areas, such as last-mile logistics, IoT drones management, wastewater monitoring, and construction. The paper by Bartoli et al. [109] formulated algorithms to position drone warehouses in delivery chains, and Guerra et al. [110] confirmed UAVs as mobile sensor networks that can monitor the environment. As was shown by Bulgakov et al. [111], drones can help cranes during the construction process and enhance their accuracy and safety. Gago et al. [112] introduced an autonomous LiDAR-based aerial platform that was able to map stockpile warehouses and estimate the volume of their materials with high accuracy, which provides significant benefits in terms of safety and efficiency compared to traditional manual approaches. Rahmadya et al. [113] came up with a validated system of calculating a safe operating distance between and RFID-tagged metallic objects to allow reliable signal readings when performing inventory operations. Likewise, Orgeira-Crespo et al. [114] and de-Dios et al. [115] contributed to the adoption of UAVs in smart factories, wherein the integration of an IoT connection and collaborative robots is used to deliver parts and locate tools. Bulgakov et al. [116] compared a coordinated robotic system in which a UAV manages a mobile crane by establishing optimal tracks to complete automated construction work. Their work demonstrates that aerial sensing coupled with Chebyshev–Gauss-based path optimization can make assembly tasks, including pergola blade positioning, fully automated. Meanwhile, Galtarossa et al. [117] and Moura et al. [118] concentrated on GPS-denied space navigation, where they used visual–inertial odometry and Graph-SLAM to ensure accurate autonomic maneuvers. These articles highlighted the versatility of drones in the logistics sector, environmental inspection, and construction, and also highlighted some of the challenges, which include battery duration and efficient localization. The RFID-SOAN system was used to design an autonomous UAV to find an indoor inventory maplessly by Alajami et al. [119]. The UAV relies on RFID tags as navigation markers, and during testing, the device had a 99-percent tag-reading success rate. Their research indicates that this is a very efficient substitute for the customary map-based navigation strategies in warehouses.

The latest contributions exhibit a trend towards multi-robot work, AI perception, and more sophisticated 3D modeling. Wang and Sherry [120] designed unified solutions to coordinate UAVs with ground vehicles and other resources, and Geetha et al. [121] and Yoon et al. [122] suggested drone systems with Wi-Fi-based positioning, RFID, and deep learning-based methods to increase the level of warehouse automation (Figure 47). Yang et al. [123] proposed UAV-UGV cooperative systems in 2024, and Tsiakas et al. [124] introduced the concept of 3D inventorying with the use of particle filters, both of which aim at enhancing coverage, localization, and safety. Pan et al. [125] went even further and introduced smart MAVs that could autonomously perform inspections and create 3D representations in high fidelity with the help of Gaussian Splatting. Collectively, these advancements reflect a direct current trend in a process wherein single-UAV prototypes evolve into distributed, multi-robot, and AI-enabled systems that will change the way industries approach inventory, logistics, and the inspection of infrastructure in increasingly dynamic and complex settings.

Advantages: Drones have several benefits when it comes to optimizing warehouse layout.

• Rapid mapping and geometry capture for digital twins: UAVs equipped with vision/LiDAR can generate 3D maps faster than manual surveying, enabling frequent layout assessment and iterative optimization.
• Operational insight through congestion and utilization observation: Repeated aerial scans can reveal bottlenecks (traffic hotspots, underutilized zones) and support evidence-based layout redesign rather than static assumptions.
• Change detection over time: UAV mapping supports monitoring of layout drift (racks moved, temporary storage growth) and helps maintain accurate facility models for planning.
• Improved decision support for slotting and routing: Spatial models can be used to redesign slotting strategies and travel corridors, reducing unnecessary motion and improving throughput.
• Reduced disruption compared with manual measurement: UAV-based data collection can often be performed off-peak with minimal interference to operations.

Limitations: Drones have several challenges when used in warehouse layout optimization.

• Mapping quality depends on lighting and surface properties: Reflective surfaces, repetitive textures, and narrow corridors can degrade vision-based mapping; LiDAR improves robustness but adds cost/payload constraints.
• SLAM drift and loop-closure failures in repetitive aisles: Warehouse geometry is highly repetitive; without strong loop closure cues, SLAM can accumulate errors and degrade map accuracy.
• Endurance constraints for large-scale 3D capture: Full-facility mapping may require multiple sorties and careful mission partitioning to ensure consistent coverage.
• Safety constraints in active operations: Conducting mapping flights in dynamic conditions increases risk and may require reduced speed or restricted flight windows.
• Data processing and integration overhead: Converting raw scans into actionable layout recommendations requires analytics pipelines (digital twin generation, congestion modeling), which can be more demanding than the data collection itself.

The study of Workshop Layout Optimization (WLO) highlights evolving research trends, technological advances, and robot–environment configurations across time, as shown in Figure 48, Figure 49, Figure 50, Figure 51, Figure 52, Figure 53 and Figure 54 and

Table 31. Detailed literature survey on the application of drone technology in warehouse layout optimization.

Ref. No.	Year	Function of Drone	Robot Type	Single/Multiple Drones	Simulation Results	Real-Time Results	Technology/ Methodology Used	Environment
106	2014	Warehouse layout optimization	UAV	Multiple	Yes	No	MAV system	Static
107	2015	Warehouse layout optimization	UAV	Single	Yes	No	Genetic algorithm	Static
108	2016	Warehouse layout optimization	UAV	Single	Yes	No	PTAM algorithm	Static
109	2018	Warehouse layout optimization	Wheeled robot	Single	Yes	No	Distributed optimization using primal–dual decomposition	Static
110	2019	Warehouse layout optimization	UAV	Multiple	Yes	No	Distributed control strategy for multi-quadrotor UAVs	Dynamic
111	2019	Warehouse layout optimization	UAV	Single	Yes	No	Exact and approximate algorithms	Dynamic
112	2019	Warehouse layout optimization	UAV	Single	Yes	No	Drone-based IoT device management platform	Static
113	2019	Warehouse layout optimization	UAV	Single	Yes	Yes	Multi-UAV network, sampling, localization	Dynamic
114	2019	Warehouse layout optimization	UAV	Single	Yes	No	Cyber–physical system and odometric approach	Dynamic
115	2020	Warehouse layout optimization	UAV	Single	Yes	Yes	GPS-based and visual/LiDAR-based localization	Dynamic
116	2020	Warehouse layout optimization	UAV	Single	Yes	Yes	RFID reader	Dynamic
117	2020	Warehouse layout optimization	UAV	Single	Yes	No	RFID, sonar sensors, artificial vision, industrial IoT	Static
118	2020	Warehouse layout optimization	UAV + wheeled robot	Multiple	Yes	No	Chebyshev–Gauss collocation method	Dynamic
119	2020	Warehouse layout optimization	UAV	Single	Yes	Yes	Aerial robot co-worker system	Static
120	2020	Warehouse layout optimization	UAV	Single	Yes	Yes	Visual–inertial algorithms and SLAM	Dynamic
121	2021	Warehouse layout optimization	UAV	Single	Yes	No	Graph-SLAM approach	Static
122	2022	Warehouse layout optimization	UAV	Single	Yes	Yes	RFID-SOAN	Static
123	2022	Warehouse layout optimization	UAV	Single	Yes	No	LAIDAVAMS (integrated drone and asset management system)	Static
124	2023	Warehouse layout optimization	UAV	Single	Yes	No	Wi-Fi-based positioning and RFID	Dynamic
125	2023	Warehouse layout optimization	UAV	Single	Yes	Yes	Deep learning with video frame continuity for 3D estimation	Static
126	2024	Warehouse layout optimization	UAV	Multiple	Yes	Yes	LiDAR–inertial odometry and target-based relative localization	Static
127	2024	Warehouse layout optimization	UAV	Single	Yes	No	Particle filter using distance sensors and IMU	Static
128	2024	Warehouse layout optimization	UAV	Single	Yes	No	Smart micro aerial vehicles (MAVs)	Static

,

Table 32. Year-wise distribution of research on static and dynamic environments in WLO.

Year Range	Static Environment	Dynamic Environment	Total
2014–2015	2	0	2
2016–2018	2	0	2
2019–2020	3	7	10
2021–2022	3	0	3
2023–2024	5	1	6
Total	14	8	23

,

Table 33. Robot–environment configuration analysis in WLO.

Robot Type	Static Environment	Dynamic Environment	Total
UAV	13	8	21
Wheeled robot	1	0	1
UAV + wheeled hybrid	0	1	1
Total	14	9	23

,

Table 34. Technology adoption in different environments in WLO.

Technology/Methodology Used	Static Environment	Dynamic Environment	Total
Genetic algorithm	1	0	1
PTAM	1	0	1
Distributed optimization (primal/dual)	1	0	1
IoT device management	1	0	1
Aerial co-worker system	1	0	1
Graph-SLAM	1	0	1
RFID-SOAN	1	0	1
LAIDAVAMS	1	0	1
Deep learning vision	1	0	1
LiDAR–inertial odometry	1	0	1
Particle filter with IMU	1	0	1
MAV/smart MAV	1	0	1
Distributed multi-UAV control	0	1	1
Exact and approximation algorithms	0	1	1
Multi-UAV network, sampling, localization	0	1	1
Cyber–physical system with odometry	0	1	1
GPS-based or visual–LiDAR localization	0	1	1
RFID reader	0	1	1
Chebyshev–Gauss collocation	0	1	1
Visual–inertial SLAM	0	1	1
Wi-Fi positioning with RFID	0	1	1
Total	14	9	23

,

Table 35. Real-world and virtual implementation analysis in WLO.

Environment	Simulation Only	Real-Time Results	Both (Simulation + Real-Time)	Total
Static	11	3	0	14
Dynamic	4	5	0	9
Total	15	8	0	23

,

Table 36. Chronological technology adoption in WLO.

Period/Phase	Dominant Technologies	Emerging Technologies
2014–2016	MAV systems, basic UAV navigation, PTAM	Optimization techniques
2018–2019	Distributed optimization, IoT platforms	Multi-UAV coordination, network sampling
2020 (peak activity)	RFID with sensors, cyber–physical systems, visual–LiDAR, Graph-SLAM	Chebyshev–Gauss approximation, hybrid UAV–UGV
2021–2022	RFID-SOAN, SLAM enhancements	Asset management automation (LAIDAVAMS)
2023–2024	Deep learning for 3D estimation, LiDAR–inertial odometry	Smart MAVs, real-time localization

and

Table 37. Functional capabilities by technology type in WLO.

Function	Primary Technologies	Accuracy Metrics	Real-Time Performance
Localization	SLAM, GPS–LiDAR, RFID-SOAN, Wi-Fi + RFID	Centimeter-level precision, multi-sensor fusion accuracy	Good performance under dynamic constraints
Path planning and control	Chebyshev–Gauss methods, distributed multi-UAV control, optimization techniques	Smooth trajectory generation, reduced jerk	Real-time feasible
Inventory scanning and tracking	Deep learning-based vision, RFID, IoT sensors	Frame continuity and detection accuracy	Moderate to high
Mapping and warehouse navigation	PTAM, Graph-SLAM, visual–inertial methods	Map consistency, drift reduction	Strong performance in structured layouts
Swarm coordination	Multi-UAV strategies, sampling and localization	Collision avoidance reliability	Requires motion capture or external anchors

. UAVs dominate as the primary robotic platform used (Figure 48), with static environments being explored more frequently than dynamic ones; 14 out of 23 studies focused on static settings (Figure 49; Table 32 and Table 33). Interestingly, research peaked in 2020 and has diversified since, which is reflected in the varied adoption of technologies such as SLAM, RFID-SOAN, distributed optimization, and deep learning vision techniques (Figure 50 and Figure 53; Table 34, Table 35 and Table 36). Although simulation-based experimentation remains more prevalent, especially for static environments (Figure 52; Table 35), there has been a notable rise in real-time implementation in dynamic contexts. Single-UAV systems are more widely adopted compared to multiple-UAV systems (Figure 51). Over the years, technology adoption has progressed from basic MAV systems to more advanced methods like lidar–inertial odometry and deep learning for 3D estimation (Figure 54; Table 36). These innovations have enabled key capabilities such as accurate localization, swarm coordination, and smooth path planning, all critical for effective inventory tracking and warehouse navigation (Table 37).

5. Results and Discussions

The main differences between conventional warehouse management systems and drone-integrated systems are efficiency, accuracy, scalability, and safety. Conventional systems predominantly utilize manual labor for activities like inspection, auditing, and data capture, which can be time-consuming, error-prone, and not efficient for large-scale operations. Drones, on the other hand, use advanced sensor technologies like LiDAR and RFID, which provide speed and accuracy in large-scale areas. Traditional methods of dealing with emergencies involve human involvement, which can take time and expose them to risks, whereas drones can safely access hazardous areas and present real-time visual feedback. Scalability is another major difference, as conventional systems require heavy investments in terms of infrastructure and manpower for scaling, whereas scaling in drones is simple with limited additional resources. Drones can also be integrated with ease into warehouse management systems, Internet of Things platforms, and analytics solutions, providing real-time insights and optimization, whereas traditional systems provide periodical updates with human intervention. Although drones involve an increased upfront cost, they save operational costs in the long term as well as improve the overall efficiency of the warehouse. They make for an ideal choice for modern warehouses with advanced technological support.

Figure 55 shows that from 2002 until 2010, research activity was very low, amounting to only 1–3% annually. From 2014 to 2017, modest growth was observed, with contributions reaching up to 4% in 2016. A sharp growth phase started in 2018 with 8%; from 2019 through 2022, consistently high annual contributions were seen at about 13% per year. With a slight setback in 2023 to 8%, the trend reached its peak in 2024 at 15%, confirming that this series is on an upward trajectory and that there has been sustained research interest in recent years.

Figure 56 shows that the research focus on inventory management dominates, accounting for 52% of the total studies. It is followed by optimization of warehouse layout (19%) and security and surveillance (13%), showing the growing interest in efficiency and safety-related applications. Picking and packing or goods transportation represents a moderate share of 10%, and maintenance and inspection remains the least-explored area at 5%. Overall, the trend underlines a strong concentration of research into core inventory operations, with comparably limited attention to support and maintenance-oriented applications.

Figure 57 indicates that, across various application areas, single-agent approaches dominate. Inventory management is found to be the most concentrated area, with 43% of research being in single-agent systems and 9% in multi-agent systems. Similarly, in warehouse layout optimization, the proportion of studies using single-agent methods is higher, at 16%, than multi-agent approaches, at 3%. Regarding security and surveillance, single-agent studies take the lead, making up 10%, compared to the 3% that are multi-agent studies. Picking and packing or goods transportation shows that the extent of research is still evenly distributed between single-agent and multi-agent systems, with 5% each, thus showing an early exploration of collaborative approaches. Very few studies are conducted on maintenance and inspection: 3% single-agent and 2% multi-agent. The general trend shows a high inclination toward single-agent solutions; multi-agent systems are still in their early days and limited to a few applications.

Figure 58 illustrates the strong dominance of static environments throughout all application areas. Inventory management has the highest share, with 49% of studies conducted in static settings compared to only 3% in dynamic environments. The same superiority of static over dynamic environments can be observed in warehouse layout optimization, which uses static environments for 12% and dynamic ones for 8% of studies. Security and surveillance offer the highest real-world variability, with 9% having taken place in dynamic environments compared to 4% in static environments. Picking and packing or goods transportation is researched mainly under static conditions, while dynamic consideration takes place in only 3%. In maintenance and inspection, both static and dynamic environments are represented by 3% each. Overall, the trend indicates a strong research bias toward static environments, while their dynamic counterparts remain comparatively underexplored except in security-related applications.

Figure 59 shows a clear bias toward simulation-based studies for all application areas. Inventory management is predominant, with 45% of the studies being simulation-based and only 7% being experimental in nature. Warehouse layout optimization (19%) and picking and packing or goods transportation (10%) are purely simulation-driven, with no experimental studies reported. Minimum experimentation is reported regarding security and surveillance: 2% experimental against 12% simulation. Maintenance and inspection remain a little-explored area, with 4% simulation and 1% experimental studies. Thus, the trend clearly indicates that inventory management research is heavily simulation-oriented, while real-world experimental validation is still relatively scarce across applications.

In summary, drones are mainly being applied to inventory and layout-related tasks, with research activity growing rapidly since 2019. Single-drone systems are still more common, but multi-drone cooperation is gaining ground. Current studies remain focused on static and simulated conditions, which highlights the need for more work in dynamic, real-world settings to unlock their full industrial potential. A comparison of traditional versus drone-enabled warehouse management systems is shown in

Table 38. Traditional warehouse management vs. drone warehouse management.

S. No.	Parameter	Traditional Warehouse Management	Drone-Enabled Warehouse Management
1	Data collection	Semi-automated or manual; time-consuming and error-prone	Fully automated, fast, and precise using cameras, RFID, and LiDAR
2	Inventory tracking	Manual scanning or fixed systems; limited reach and coverage	Dynamic, real-time tracking with access to hard-to-reach areas
3	Layout optimization	Based on periodic audits, manual measurements, and static analysis	Continuous optimization using 3D maps, digital twins, and workflow analytics
4	Emergency response	Dependent on human availability; higher risk in hazardous zones	Immediate access to dangerous areas with live visual feedback
5	Speed of operations	Slow and labor-intensive, especially during audits	Fast and automated, covering large warehouse areas efficiently
6	Cost	High operational cost due to labor dependency and inefficiencies	Higher initial cost but reduced labor and long-term operational savings
7	Accuracy	Prone to human error in data recording and audits	High accuracy through automated sensing and data processing
8	Adaptability	Limited adaptability to dynamic inventory or workflow changes	Highly adaptable with real-time monitoring and rapid response
9	Safety	Manual operations in hazardous zones increase risk	Safer operations by eliminating human exposure to danger
10	Scalability	Scaling requires significant labor and infrastructure investment	Easily scalable with minimal infrastructure modification
11	Promotional and marketing	Limited to ground-level visuals and static imagery	High-quality aerial visuals showcasing innovation and efficiency
12	Energy efficiency	Relies on energy-intensive equipment, such as forklifts	More energy-efficient for comparable inspection and monitoring tasks
13	Technology integration	Disparate systems for inventory, inspection, and emergency tasks	Seamless integration with WMS, IoT, and analytics platforms
14	Maintenance	Manual inspection with downtime for equipment	Automated maintenance checks reduce downtime
15	Real-time insights	Delayed insights due to manual data processing	Real-time analytics enabling faster decision-making

.

➢

Increased Efficiency and Speed: Drones can be used to accomplish tasks at a much faster rate than manual systems. As an example, a physical audit of stock in a warehouse, which would otherwise require days, can be completed in a matter of hours. RFID scanners and cameras attached to drones can scan and track stock automatically and speed up operations significantly. DHL uses drones in warehouses, for example, to conduct automated stock checks and save man-hours on manual stock counts and throughputs.

➢

Improved accuracy and minimized human error: Drones are very precise in the data that they capture. They are able to read the barcodes or RFID tags of items without human intervention, which negates the chances of human error in recording stock quantities or mislocation. One example is the use of drones by Amazon to perform live stocktaking to ensure the proper tracking of stock and, subsequently, do away with costly errors that are the result of human variation.

➢

Enhanced Safety: The use of drones does not require human workers to enter areas of the warehouse that may be dangerous, like high shelves or dangerous locations. For example, Walmart employs drones to scan the upper shelves in their huge distribution centers to enable them to monitor their stock without risking the safety of their employees through falling or injuries they may incur due to forklifts.

➢

Labor and Operation Cost Savings: Drones require a certain initial investment, but in the long run will save the cost of manual work since they are able to replace some of the work performed by humans, such as tracking inventories, monitoring, and inspection, among others. UPS has used drones to sort packages, which reduces its reliance on human employees to carry out routine tasks. This saves on operational costs in the long run since fewer employees will be used to do the jobs that can be handled effectively by drones.

➢

Real-time Data and Analytics: Drones with cameras, temperature sensors, and LiDAR can supply real-time visual and spatial data to WMS, providing managers with immediate insights into the performance of their warehouses. As an example, Best Buy uses drones to track live information from its distribution centers so that it can optimize the amount of stock, as well as the design of warehouses, to improve its stock management and use of space.

➢

Improved Warehouse Layout Optimization: Drones have the potential to constantly scan and map warehouses so that their layouts are optimized to ensure better flow and productivity. For example, Zara uses drones in its warehouses to track real-time product locations and also track the workflow; hence, it can continually improve its layout to be as efficient as possible.

➢

Scalability and Flexibility: Drone systems can be modified to suit the needs of a warehouse with minimal adjustment to the existing infrastructure. It is possible to start with a few drones in a small warehouse, and, as the operations grow, more drones can be introduced. This is what FedEx did in the case of drones in its warehouses as a response to the increased number of packages in e-commerce, which required minimal infrastructure changes.

➢

Advanced Emergency Response: Drones can be flown very quickly in case of an emergency to check what is going on, be it a fire, a gas leak, or an injured employee. The presence of thermal imaging drones will allow the detection of hotspots and ensure the safety of the workers before an inspection. An example here is Google Wing, which uses drones in emergency response, where human presence would be dangerous, to allow first responders to have much-needed real-time information in an emergency or a disaster.

➢

Sustainability: Drones are also energy-efficient in comparison to traditional equipment used in warehouses, such as forklifts, which saves on power consumption. Drones can contribute to eliminating the carbon footprint of warehouses by reducing the use of systems that burn fuel. L’Oréal uses drones in its warehouses to automate work, hence minimizing its dependence on manual labor as well as power-guzzling machinery, allowing the company to achieve sustainability targets.

Drone-based warehouses dramatically enhance operational efficiency, save costs, improve accuracy, and boost safety. With examples from such notable businesses as Amazon, DHL, and Walmart, drones have emerged as a game-changer in contemporary warehouses, helping businesses stay competitive in rapidly changing, technologically advanced sectors. With further advancements in the use of drones, their advantages in warehouses will only grow, solidifying their position as an indispensable tool for logistics and supply chain management in the future.

5.1. Challenges of Drone-Based Warehouse Solutions

Drone-based warehousing systems, as efficient as they are precise, also have several drawbacks.

▪

One is the high initial cost of drones, sensors, software, and the upgrading of infrastructure, which is a hindrance to small businesses.

▪

Another is that one has to go through complicated regulatory entities, adhere to air traffic and privacy laws, and thereafter find the required permits, which can be costly and time-intensive.

▪

Drones also need to be regularly serviced and maintained, which only skilled repairmen can provide, which in turn adds to their long-term costs. It is also necessary to train the existing workforce, which makes it expensive in terms of raising man-hour costs, along with low scalability.

▪

Drones have a small payload and cannot handle heavy objects; they need to be supported by conventional equipment.

▪

Weather conditions, including rainfall, wind, or fog, may compromise the functionality of drones, especially in open spaces or outdoors.

▪

There is also the technical aspect of adopting drones in conjunction with the existing warehousing systems; this requires a lot of IT assistance.

▪

Security is also a concern, as drones are vulnerable to cyberattacks that can cause system shutdown or hacked facilities.

▪

Resistance of employees because of the perceived violation of their privacy, and the limited battery capacity and range of drones may affect both adoption and normal usage.

Overcoming all these issues will take strategic investment, knowledge of legislation, development of workforce skill sets, and transparency of communication with stakeholders.

5.2. Determining Literature Gaps

Despite notable advancements in the use of drones in the management of warehouses, several knowledge gaps must be filled for their effective integration.

▪

Not much is known regarding the long-term impact on the workforce in terms of job displacement, role change, and reskilling requirements.

▪

There is also a lack of information regarding successful human–drone cooperation, such as the ability to build trust, user interfaces, and safe coordination.

▪

Privacy and ethical concerns, such as how to keep the data of drones private as well as the problem of workers being monitored by them, are poorly studied and require more definite rules.

▪

Although drones are viewed as sustainable, there is no empirical evidence on the power efficiency of drones versus conventional equipment and their impact on the environment.

▪

Technical challenges, e.g., the connection of drones with outdated warehouse systems, also require additional research.

▪

Most papers assume large-scale operations, not paying attention to cost feasibility for small- and medium-sized warehouses.

▪

Health and safety concerns, including noise and risk of injury, as well as physical impact, have not been sufficiently researched.

▪

Studies are also necessary on enabling real-time decision-making with AI and machine learning, as well as multiple-drone management.

▪

Regulatory advancement and ensuring that drone systems comply with legal documents must be continually evaluated.

▪

Combining warehouse drones with last-mile delivery is an undeveloped avenue of research.

Filling these gaps is important for increasing drones’ usefulness in warehouses and for ensuring responsible, scalable, and effective deployment.

5.3. Future Prospects

The outlook for warehouse drones is promising, and they have a transformational capability to change the supply chain as well as the warehouse process. With the development of drone technology, it will become more closely involved with other automated systems, such as robots and IoT devices, becoming more efficient. Drones will be autonomously navigated and make real-time decisions with the help of AI and machine learning, and will be developed to fulfill many other functions, including maintenance and monitoring. Their coordination and perception will improve with the use of swarm robotics and sophisticated sensors. New technologies such as edge computing, enhanced battery life, and digital twin simulations will make productivity and planning more productive. The drones can also be integrated with last-mile delivery systems, which will increase the speed and flexibility of logistics. Given the changing regulations, drones will assist warehouses in becoming more efficient and sustainable.

5.4. Industrial Case Studies and Recent Real-World Deployments

Recent evidence shows that drone-enabled warehouse management is transitioning from lab validation to scaled industrial operation [127]. For example, Ingka Group (IKEA) and Verity report large-scale autonomous inventory deployments across international warehouse networks, including over 250 autonomous drones in 73 warehouses and daily scanning volumes above one million pallets [IKEA/Ingka Group]. Additional industry reports describe multi-site enterprise adoption (e.g., Gather AI deployments linked to GEODIS operations) and notable operational improvements in cycle counting and productivity [IKEA/Ingka Group]. Complementing these deployments, recent research demonstrates practical warehouse validation of full, 3D UAV inventory workflows, including strong detection/counting performance under realistic conditions, while earlier work established architecture-level proof-of-concept for mini-drone inventory automation [127].

6. Conclusions

The review has conducted a critical and methodical synthesis of more than 120 research articles on drone-facilitated warehouse practices, giving a comprehensive overview of the application of unmanned aerial systems in contemporary logistic settings. The current study brings together divergent research efforts through a systematic investigation of the functions of drones, robotic platforms, deployment architectures, validation strategies, and enabling technology, as well as environmental contexts. The analysis revealed that warehouse drones are highly promising in transforming inventory management, inspection, surveillance, and intralogistics with respect to improving their operational flexibility, accuracy, and efficiency, particularly in large-scale and/or densely stacked vertical warehouses. The majority of the currently available literature is confined to single-UAV systems and can be proven with simulations; not many have been proven to be autonomous and operational in real time in a warehouse setting. Although multi-drone systems are becoming a more popular research topic to address the issues of scalability and efficiency, the challenges of coordination, collision avoidance, communication reliability, and safety remain significant barriers to realistic implementation. In addition, the intense preoccupation with inanimate surroundings highlights a lack of connection between research assumptions and real-life warehouse circumstances, dynamic challenges, human behaviors, and dynamic layouts.

A major portion (52%) of existing research focuses on inventory management, highlighting it as the primary area of interest. In contrast, maintenance and inspection receive minimal attention (5%), revealing a significant research gap in support functions. This imbalance underscores the need for broader exploration beyond core operations to enhance overall warehouse efficiency. Overall, the present review not only summarizes the state of the research on drone and warehouse interaction but also establishes clear directions for further work. It will be essential to look further into multi-drone coordination, strong safety-conscious autonomy, real-time validation, and the ability of drones to operate in dynamic settings to translate laboratory work into industrial use. The organized tables and comparisons in this work will be a valuable resource for researchers and practitioners alike, and will assist in making relevant decisions and moving future research toward more scalable, reliable, and industry-ready drone-based warehouse systems.