Design and Realization of an Intelligent Production Line for Particle-Containing Bottled Product ^†

Abstract

The research explored the automation production lines for the bottling of particulate materials in the pharmaceutical industries, covering the integrated processes of loading bottles, filling with particles, sealing, screwing on caps, quality inspection, and storage. The hardware system of the project consists of programmable logic controllers(PLCs), edge servers, motion control equipment, industrial cameras, and mechanical grippers for handling and storage. The aim of this research is to assist the manufacturing industry in transitioning from traditional production models to digital and intelligent production methods. From the perspective of core components, it analyzed and expounded the key technologies for building a digital production line; at the same time, from the perspective of data collection and processing, it clarified the role and advantages of the cloud platform. The product packaging process simulation covers loading bottles, filling with particle materials, sealing, screwing on caps, quality inspection, and storage. The production line issues production instructions and scheduling plans through the human-machine interaction interface and the cloud platform.

1. Introduction

The manufacturing landscape is undergoing a profound transformation driven by the imperatives of Industry 4.0. In the pharmaceutical sectors, where product safety, traceability, and packaging precision are paramount, the shift from traditional, labor-intensive production lines to intelligent, automated systems is not merely an option but a necessity. Bottled products containing granular materials—such as pharmaceutical pellets, nutritional supplements, or other granules—present unique challenges. Traditional production lines often struggle with issues like inconsistent fill volumes, difficulties in real-time quality monitoring, lack of predictive maintenance, and inefficient material handling, leading to reduced productivity and increased waste.

2. System Architecture and Core Technologies

The intelligent production line elaborates on the system’s design and realization from three core technological dimensions: the hierarchical control architecture centered on the PLC, the robust industrial networking that enables device interoperability, and the application of digital twin technology for virtual commissioning and real-time monitoring. Furthermore, it discusses the pivotal role of the cloud platform in aggregating production data, enabling informed decision-making, and facilitating remote scheduling. The objective is to provide a viable and scalable model for the digital and intelligent transformation of manufacturing enterprises specializing in granular product bottling.

The intelligent production line is designed as a modular and scalable system, with its hardware architecture forming the foundation for all automated operations. The physical setup is a linear arrangement of interconnected modules, synchronized to perform the sequential steps of the bottling process.

2.1. Core Mechanical and Control Hardware

The primary hardware components include PLC, Motion Control Devices, Industrial Cameras and Vision Systems, Robotic Manipulators, Mechanical Grippers, and Edge Servers.

Programmable Logic Controllers (PLC): Serving as the central nervous system, a high-performance PLC (e.g., Siemens S7-1500 series) is responsible for executing the master control program [1]. It coordinates the sequence of operations, manages I/O devices, and ensures precise timing between modules. Safety PLCs are integrated to manage emergency stops and safety light curtains. Motion Control devices: High-precision servo drives and motors are employed to control the indexing of the conveyor system, the rotational speed of the filling turret, and the torque applied during the screw capping process [2]. Industrial Cameras and Vision Systems: High-resolution, high-speed cameras are positioned at critical inspection stations [3]. They capture images of filled bottles to verify fill levels, detect foreign particles, and inspect cap placement and seal integrity. Robotic Manipulators and Mechanical Grippers: For handling and storage, articulated robots equipped with custom-designed mechanical grippers are used [4,5]. These grippers are engineered to gently but securely handle bottles and primary packaging materials without causing damage. Edge Servers: Located on the factory floor, edge servers act as the bridge between the physical hardware and the cloud [6]. They aggregate raw data from the PLCs and vision systems, perform real-time data filtering and analysis, and host lightweight digital twin models for immediate local feedback.

2.2. Sensor Network and Actuator

A dense network of sensors (inductive, capacitive, photoelectric, and vision-based) provides real-time feedback on the status of every component [7]. These sensors detect bottle presence, monitor fill levels, confirm cap application, and track the position of robotic arms. Actuators, including pneumatic cylinders and electric drives, execute the physical movements commanded by the PLC, such as rejecting a faulty bottle or diverting a batch to the warehousing area.

2.3. Key Technologies for Intelligence Production

The intelligence of the production line arises from the synergistic integration of advanced control, communication, and simulation technologies [8].

2.3.1. PLC Master Control and Equipment Networking

The software logic within the PLC is the heart of the production process. It is programmed to handle complex sequences, fault diagnosis, and interlock conditions [1]. The control program manages Process Synchronization, Granule Filling Control, and Quality Inspection Logic.

Process Synchronization: Ensuring that bottles are fed, filled, and capped at the precise rate to avoid bottlenecks or jams [9]. Granule Filling Control: Precisely controlling the filling mechanism (e.g., auger filler or volumetric cup) to dispense the exact weight or volume of granules into each bottle [2]. Quality Inspection Logic: Receiving signals from the vision system and triggering the rejection mechanism for any non-conforming product [3,4].

Equipment networking is achieved through industrial Ethernet protocols (e.g., ProfNet, Ethernet/IP). This network connects the PLC, HMI, edge servers, vision systems, and drives, creating a unified, high-speed communication backbone. This allows for seamless data exchange and centralized monitoring from the HMI [1].

2.3.2. Data Collection and Cloud Platform Integration

The edge server plays a crucial role in data collection [6]. It continuously streams data from the production line—operational status, production counts, alarm events, and quality metrics—to a cloud-based platform. The cloud platform serves multiple functions: Centralized Data Aggregation, Big Data Analytics, Production Order and Schedule Management.

Centralized Data Aggregation: It consolidates data from multiple production lines across different facilities, providing a holistic view of manufacturing operations [6]. Big Data Analytics: Historical data is analyzed to identify trends, predict potential equipment failures, and optimize overall equipment effectiveness (OEE).Production Order and Schedule Management: The cloud platform acts as the system of record for production planning. Orders and schedules are created here and then pushed down to the line’s HMI for execution, closing the loop between enterprise resource planning (ERP) and shop-floor execution [9].

2.3.3. Digital Twin and Virtual Simulation

The digital twin of the entire production line is constructed using 3D modeling software [6,8]. This virtual replica is more than just a visual simulation; it is dynamically linked to the real-time data from the physical line. The digital twin is used for Virtual Commissioning, Real-Time Monitoring and Diagnostics, and Operator Training.

Virtual Commissioning: Before any physical hardware is built, the control logic can be tested and validated within the virtual environment, drastically reducing commissioning time and cost [8].Real-Time Monitoring and Diagnostics: The 3D model mirrors the actual line in real-time, allowing operators and engineers to visualize the production flow, identify the location of a fault instantly, and simulate the impact of a process change [6]. Operator Training: It provides a safe and realistic environment for training new operators on machine operation and troubleshooting procedures.

3. System Realization

3.1. Production Order Management

The production process begins with order creation and release through the cloud-based MES [9]. The order management system handles Order Creation, Material Availability Check, Order Scheduling, and Order Release.

Order Creation: Orders can be created manually through the MES interface or automatically generated from ERP sales orders. Each order specifies: Product type and formulation (including granule characteristics), Batch size and lot number, Packaging materials (bottle type, cap type, label specifications), Quality sampling requirements, Due date, and priority. Material Availability Check: Before order release, the system verifies the availability of Bulk granules in the appropriate grade and quantity, Empty bottles in the correct specification, Caps and closures, Labels and packaging materials. Order Scheduling: The scheduling algorithm considers the current line status and scheduled maintenance, Changeover times between products, Material availability and delivery schedules, Order priorities and due dates [9]. Order Release: Approved orders are transmitted to the line HMI, where the operator can review and acknowledge the order before initiating production.

3.2. Process Workflow Description

• The complete production workflow encompasses the following sequential operations.

Stage 1: Bottle Feeding and Singulation

Empty bottles are delivered to the line in bulk containers or pallets. A bottle unscrambler orients and feeds bottles onto the conveyor in a single-file stream. Key features include a hopper with level control to maintain a consistent bottle supply, a centrifugal or vibratory bowl feeder for bottle orientation, a reject mechanism for damaged or misoriented bottles, and backlog control to prevent bottle jams [9].

Stage 2: Bottle Cleaning and Sterilization (Pharmaceutical Applications)

For pharmaceutical applications, bottles pass through a cleaning station: Ionized air rinsing to remove particulate contamination, UV sterilization for microbial control, Vacuum extraction to remove dislodged contaminants, HEPA-filtered air shower to maintain cleanliness.

Stage 3: Granule Filling

The filling station employs a servo-driven auger filler with gravimetric verification [2]: Bulk granules are stored in a supply hopper with level control, An auger feed system conveys granules to the filling nozzle, Each bottle is indexed under the filling nozzle, The auger rotates a precise number of revolutions to dispense the target fill weight, A checkweigher immediately after filling verifies fill weight, Out-of-tolerance bottles are automatically rejected, Fill parameters are adjusted in real-time based on checkweigher feedback, For products requiring precise fill weight control, a two-stage filling approach is employed: Fast bulk fill to approximately 90% of target, Slow dribble fill to final target weight, This approach maximizes speed while maintaining accuracy.

Stage 4: Cap Placement

Filled bottles proceed to the capping station: Caps are fed from a bulk hopper through a cap sorter and elevator. A pick-and-place unit picks each cap and places it loosely on the bottle, Cap presence and orientation are verified by a vision system [3,4]. A reject mechanism removes bottles with missing or misoriented caps.

Stage 5: Screw Capping

The screw capping station tightens caps to a specified torque: Bottles are held securely during capping to prevent rotation, Spindles with torque-controlled drives engage the caps, Capping torque is monitored continuously and recorded for each bottle, Statistical process control charts track torque consistency, Bottles with torque outside specification are rejected. For child-resistant or tamper-evident closures, additional verification steps ensure proper engagement of the safety feature.

Stage 6: Induction Sealing (Optional)

For products requiring hermetic seals, an induction sealing station applies a foil seal to the bottle mouth, uses electromagnetic induction to bond the seal to the bottle, and verifies seal integrity through visual inspection or pressure testing.

Stage 7: Labeling and Coding

Labels are applied, and codes are printed: Labels are applied with precision registration. Printers apply lot numbers, expiration dates, and barcodes. Vision systems verify label presence, position, and print quality [3], 2D data matrix codes are verified for readability and data accuracy.

Stage 8: Final Quality Inspection

Comprehensive inspection verifies all quality attributes: Fill level verification, Cap torque verification (audit sampling), Label inspection, Seal integrity verification, Foreign particle detection (vision or X-ray inspection) [3,4], Statistical sampling for off-line testing.

Stage 9: Accumulation and Buffering

An accumulation table provides buffer capacity between the high-speed filling line and slower packaging operations: Bottles accumulate when downstream equipment is temporarily stopped, Sensors monitor the accumulation level and control the upstream speed [7]. This decoupling improves overall line efficiency.

Stage 10: Case Packing and Palletizing

Finished bottles are prepared for shipment: Bottles are grouped into pack patterns, Case erectors form shipping cases, Robots load bottles into cases [5], Case sealers apply adhesive and close cases, Palletizing robots stack cases onto pallets [9], Stretch wrappers secure pallet loads, and labeling applies shipping labels and barcodes.

Stage 11: Warehousing Integration

Completed pallets are transferred to the automated warehouse: Conveyors transport pallets to storage locations, the warehouse management system assigns storage positions [9], Automated cranes or shuttles retrieve pallets for shipping, and real-time inventory tracking maintains accuracy.

The complete workflow was simulated using discrete event simulation software (Siemens Plant Simulation) to validate performance before physical implementation [6,10]. Key simulation results include: Throughput Analysis: The simulation predicted a sustained throughput of 65 bottles per minute, with a peak capacity of 72 bottles per minute achievable during optimal conditions. Bottlenecks were identified at the filling station and capping station, guiding design optimization. Buffer equirements: Simulation determined optimal buffer sizes between stations to maintain continuous operation despite minor stoppages. Results showed that a 50-bottle buffer between filling and capping reduces efficiency loss from 8% to 2% [9].Changeover Analysis: Changeover time between products was simulated at 45 min for major changeovers (different bottle size) and 15 min for minor changeovers (same bottle, different granule). These times meet the target for flexible manufacturing. Reliability Analysis: Using historical failure data for similar equipment, the simulation predicted overall equipment effectiveness (OEE) of 82%, with availability of 88%, performance of 95%, and quality of 98% [10].

4. Conclusions

This study designed and implemented an intelligent automated production line specifically for bottled products containing granular materials in the pharmaceutical industry. From a technological perspective, this study achieves production line intelligence through three key dimensions: First, the hierarchical control architecture based on PLCs ensures precise synchronization and coordinated operation among stations; second, industrial Ethernet protocols enable seamless device interoperability, laying the foundation for data collection and centralized monitoring; third, the application of digital twin technology enables virtual commissioning, significantly reducing system deployment time while facilitating real-time monitoring and diagnostics of the production process [8].

This research provides a viable digital transformation solution for granular material bottling in the pharmaceutical industry, effectively addressing challenges prevalent in traditional production lines, such as inconsistent filling, quality monitoring difficulties, and a lack of predictive maintenance. Through deep integration of physical systems, information systems, and digital models, the proposed intelligent production line not only enhances production efficiency and product quality but also creates a safer and more valuable work environment for operators, offering practical references for the application of intelligent manufacturing in pharmaceutical packaging.