Next Article in Journal
Optimizing Helium Exploration: Noble Gas Isotopes as Probes for Migration–Enrichment Processes
Previous Article in Journal
Evaluation System of AC/DC Strong–Weak Balance Relationship and Stability Enhancement Strategy for the Receiving-End Power Grid
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Integrating Autonomous Trucks into Human-Centric Operations: A Path to Safer and More Energy-Efficient Road Transport

Faculty of Navigation, Gdynia Maritime University, 81-225 Gdynia, Poland
*
Author to whom correspondence should be addressed.
Energies 2025, 18(16), 4219; https://doi.org/10.3390/en18164219
Submission received: 2 July 2025 / Revised: 31 July 2025 / Accepted: 4 August 2025 / Published: 8 August 2025

Abstract

The increasing integration of autonomous driving technologies into heavy-duty road transport requires a clear understanding of how these systems affect professional drivers’ working time, vehicle utilization, and regulatory compliance. This study develops a model-based comparative analysis to assess the cooperation between human drivers and autonomous trucks at SAE Levels 3 and 4. Using EU Regulation (EC) No 561/2006 as a legal framework, single-driver, double-driver, and ego vehicle scenarios were simulated to evaluate changes in working time classification and vehicle movement. The results indicate that Level 3 automation enables up to 13.25 h of daily vehicle movement while complying with working time regulations, compared with the 10-h limit for conventional operation. Level 4 automation further extends the effective movement time to 14.25 h in double-crew configurations, offering opportunities for increased efficiency without violating labor codes. The novelty of this work lies in the quantitative modeling of human–machine collaboration in professional transport under real regulatory constraints. These findings provide a foundation for regulatory updates, tachograph adaptation to AI-driven vehicles, and the design of hybrid driver roles. Future research will focus on validating these models in real-world transport operations and assessing the implications of Level 5 autonomy for logistics networks and labor markets.

1. Introduction

The profession of transporting goods or people as a professional driver involves adhering to various regulations. Becoming a professional driver involves obtaining a driver’s license and the necessary permits for transportation. However, it is a challenging job that is not highly regarded in society. According to data from the annual reports of the International Road Transport Union (IRU) in recent years, the number of people working as professional drivers in Europe has decreased, despite an increase in demand for transportation services. This article is based on the thesis that self-driving technology, in conjunction with professional drivers, has the potential to revolutionize the transportation industry. The novelty of this study lies in the quantitative modeling of human–machine collaboration in professional road transport under real regulatory constraints. Unlike previous research, which has mainly focused on the technological aspects of autonomous driving or driver perception studies, this work integrates EU working time regulations with operational scenarios for SAE Level 3 and Level 4 trucks. By simulating single-driver and double-crew configurations, the study provides a unique assessment of how self-driving capabilities influence vehicle movement time, driver workload classification, and potential transport efficiency gains. This approach delivers practical insights for regulatory updates, tachograph adaptation, and the phased adoption of autonomous heavy vehicles in logistics operations.
The growing shortage of professional truck drivers and the increasing demand for safer, more efficient, and continuous road freight transport create a strong motivation to explore the integration of autonomous driving technologies into logistics operations. While numerous studies have focused on the technical development of autonomous trucks and the perception of drivers or the public, there is a clear research gap in the operational assessment of human–machine collaboration under real regulatory constraints. In particular, existing literature rarely integrates EU working time regulations, tachograph-based activity recording, and the practical feasibility of SAE Level 3 and Level 4 automation. The main contributions of this study are threefold. First, it develops a simulation-based model of professional driver and autonomous vehicle collaboration, evaluating scenarios with single-driver and double-crew operations under SAE Level 3 and Level 4 automation. Second, it provides an operational assessment that integrates vehicle movement time, driver workload, and legal compliance with EU regulations. Third, the study highlights potential efficiency and safety benefits, offering a structured framework for the gradual adoption of autonomous trucks in professional road transport.
The main objective of this study is to evaluate how the integration of SAE Level 3 and Level 4 autonomous driving systems with professional drivers affects vehicle movement time, driver workload classification, and compliance with EU working-time regulations. The central research question is: How can human–machine collaboration in partially or highly automated heavy-duty transport improve operational efficiency while remaining compliant with regulatory requirements? The study is based on the following working hypotheses: (H1) SAE Level 3 automation increases daily vehicle movement without breaching driver working-time limits; (H2) SAE Level 4 automation allows self-driving periods to be treated as availability or rest, further extending effective transport time; (H3) hybrid human–machine operations can provide a regulatory basis for the concept of a “virtual co-driver” in future tachograph systems. This article aims to answer the question: “What will the collaboration between a professional driver and a partially or fully automated vehicle look like?” This paper displays sample workdays that professional drivers can record on a digital tachograph and identifies places where the self-driving of a heavy vehicle can impact the actual working hours of a professional driver based on their workday charts. The text explains how to record driving hours in the tachograph when the vehicle is under the driver’s control and when self-driving. It is worth noting that the models presented to showcase professional drivers’ workdays are not exact representations, but rather simplified versions that highlight the best possible outcomes. Despite this, these models demonstrate the potential for teamwork between humans and machines.

2. Literature Review

The ongoing transformation of the transportation sector—driven by automation, digitalization, and sustainability imperatives—has brought autonomous heavy vehicles to the forefront of scholarly and industrial interest. Research indicates that the integration of self-driving systems with human operators, particularly professional drivers, is not only a technological challenge but also a socio-regulatory one, requiring in-depth evaluation of safety, work patterns, and user acceptance.
Driver interaction with autonomous truck systems remains a central research focus. Fu et al. (2025) explored how drivers perceive different functional aspects of autonomous trucks in China, highlighting their expectations for improved safety and control transparency [1]. Complementing this, Bhoopalam et al. (2023) captured feedback from professional drivers in Europe, indicating concerns about job security alongside recognition of operational benefits such as reduced cognitive fatigue [2]. Human–Machine Interface (HMI) design plays a crucial role in the effectiveness and safety of semi-automated vehicles. Wang et al. (2024) provided a systematic review of HMI strategies in automated vehicles, emphasizing the need for intuitive, context-sensitive designs to support critical decision-making [3]. Similarly, Sekadakis et al. (2023) argued that seamless interaction between human and machine enhances trust and minimizes miscommunication during high-risk transitions [4]. The safety impact of automation has been evaluated from both engineering and human factors perspectives. Zhao et al. (2025) proposed ensemble modeling to predict AV system reliability, particularly in complex urban environments [5], while Wang et al. (2024) developed learning-based models for predicting human behavior and system handover in mixed-control scenarios [6].
Public and driver acceptance is pivotal for large-scale deployment. Castritiuset al. (2020) analyzed the public’s openness toward semi-autonomous truck platoons, identifying a need for transparent communication and gradual integration [7]. Kabbani et al. (2023) introduced adaptive communication systems to facilitate trucker–vehicle dialogues under critical conditions [8]. External HMI (eHMI), such as visual or auditory cues visible to other road users, has become a key design consideration. Gwak et al. (2025) demonstrated that well-timed eHMI signals significantly improve pedestrian response in interactions with Level 4 trucks [9]. Brill et al. (2023) extended this discussion to shared spaces, where intuitive eHMI improved predictability and reduced pedestrian hesitation [10]. In addition, the work environment of professional drivers is increasingly mediated by autonomous systems. Fonseca et al. (2025) provided a systematic review of driver monitoring technologies (e.g., fatigue sensors, IVMS) to maintain safety in long-haul operations [11]. Yan et al. (2024) assessed both cognitive and physiological workload under cooperative driving conditions, finding measurable stress reduction in well-integrated systems [12]. Kim et al. (2022) further confirmed that partial automation can alleviate psychological strain when designed to align with driver expectations [13].
From a regulatory and Operational Design Domain (ODD) standpoint, Kim et al. (2024) evaluated safety metrics for various automation levels, suggesting that well-defined ODD constraints contribute to safer transitions between control modes [14]. Guo et al. (2022) [15] and Alhawiti et al. (2024) [16] addressed usability and real-time adaptability of eHMI, reinforcing their potential to enhance traffic harmony in mixed environments. Several studies also focused on auditory feedback and multimodal interaction. Agredo-Delgado et al. (2022) [17] and Zheng et al. (2022) [18] demonstrated that combining visual and auditory cues facilitates quicker response times during takeover requests, enhancing system transparency and reducing anxiety. Moreover, fatigue and distraction remain critical issues. Visconti et al. (2025) reviewed state-of-the-art monitoring systems that detect drowsiness and inattention in commercial drivers, recommending integration with autonomous driving platforms [19]. This aligns with the broader goal of ensuring redundancy and resilience in both human and machine components of driving systems.
In addition to technological advancements and regulatory frameworks, the operational reliability of autonomous trucks is strongly affected by external factors related to safety, security, and environmental conditions. Adverse weather such as heavy rain, snow, or fog can significantly degrade the performance of perception systems, including cameras, lidar, and radar [20], thereby limiting the effectiveness of the Object and Event Detection and Response (OEDR) function [21]. Security vulnerabilities are equally critical; autonomous trucks rely on sensors and communication modules that may be disrupted by electromagnetic interference from environmental sources such as nearby signal towers, or even by intentional electromagnetic attacks, which can lead to visual recognition errors or temporary failures in power electronics [22]. Furthermore, autonomous vehicles remain susceptible to cyber threats, including GPS spoofing and denial-of-service attacks [23], which can affect vehicle control and sensor reliability [24].
To mitigate these challenges, human-aided strategies play a vital role, particularly in SAE Level 3 and conditional Level 4 scenarios, where the presence of a professional driver allows immediate intervention in the event of sensor failure, adverse weather, or suspected cyber interference. Integrating these considerations into the design and operation of semi-autonomous trucks enhances overall resilience and supports the safe, gradual adoption of higher levels of vehicle automation.
Recent research on autonomous truck platoons has highlighted their potential to enhance both energy efficiency and freight transport capacity by reducing aerodynamic drag and enabling coordinated vehicle movement. Advanced modeling approaches increasingly leverage graph neural networks (GNNs) to capture the complex aerodynamic interactions within platoons operating under heterogeneous traffic conditions. For example, Wang et al. developed a GNN-based framework for aerodynamic analysis of truck platoons, demonstrating that traffic heterogeneity and inter-vehicle spacing significantly influence drag reduction and stability [25]. Similarly, Chen et al. proposed neural network surrogate models for aerodynamic analysis in truck platoons, providing insights into fuel efficiency improvements and operational strategies for autonomous freight delivery [26]. These studies underline the growing importance of data-driven aerodynamic modeling as a component of autonomous logistics systems and provide a complementary perspective to the operational and regulatory focus of the present study.
The reviewed literature suggests that while the path toward autonomous heavy vehicle integration is technologically feasible, its success heavily relies on driver inclusion, regulatory frameworks, and adaptive design. As the boundary between human and automated control becomes increasingly dynamic, researchers and developers must focus not only on system efficiency but also on human-centered performance indicators.
In summary, prior research has not yet integrated EU working-time regulation with SAE L3/L4 operations in a way that yields auditable, tachograph-consistent duty cycles; it rarely quantifies how TOR/MRM and driver-availability constraints reshape daily movement time for single and double crews; and it lacks a practical path toward logging automated operation (e.g., a “virtual co-driver”). Our study addresses these gaps by constructing a regulation-constrained model of human–machine collaboration that (i) assigns automated periods to legally defined activity categories, (ii) quantifies movement-time effects across L3/L4 scenarios, and (iii) embeds handover and monitoring assumptions that can be stress-tested in future sensitivity analyses.

3. Technological Context

The concept of self-driving vehicles was born in the minds of visionaries in the first half of the last century. However, the transfer of human dreams and wishes to real vehicles came much later. In Europe, in 1986, at the Hochschule der Bundeswehr München, work was carried out on the creation of a semi-autonomous vehicle. The van was able to navigate the streets (without other vehicles), controlling its lateral position and reacting correctly up to a speed of 96 km/h in terms of longitudinal control of the vehicle position. Then, for the first time, control software based on data obtained from sensors was used [27]. The team of people working on this project influenced the development of the “Eureka Prometheus” project in the early 1990s. The project involved 50 vehicles that were in normal traffic on the Munich–Copenhagen route, and the average human participation at each intersection was around 9 km [28]. At the same time, in the USA, Carnegie Mellon University designed and built a series of NAVLAB vehicles. Between 1986 and 2000, work on self-driving vehicles continued to develop. A 1996 version of the NAVLAB vehicle competed in the “No Hands Across America” event. The route covered a distance of 4800 km at a maximum permissible autonomous speed of 110 km/h, and 98.2% of the planned route was completed without human intervention [29]. In 1992, the concept of platooning travel was born. The easiest way to compare these concepts is with dog sledding, where the first dominant dog leads the others. Guidance in early platooning projects was based on magnetic markers and supported by radar, which was used for the first time, as described by Tsugawa et al. [30]. Remote communication between vehicles was also implemented, which, according to Kavathekar and Chen [31], laid the foundation for the development of modern cruise control and emergency braking systems. Another booster in the subject of autonomy was the DARPA (Defense Advanced Research Projects Agency) competition launched in 2002, which, in 2007, was transformed into the DARPA Urban Challenge. Substantial financial rewards were offered for completing the challenge ($1 million in 2004). The challenge first relied on 100% autonomous passage of the route, initially through the desert, and, after meeting this requirement, along the urban section [32]. In 2005, after several years of running without a winner, the goal of crossing the desert was achieved. Five crews crossed the desert on their own, and the winner of the competition was Stanford University’s vehicle, Stanley. The 2007 DARPA Urban Challenge was won by Carnegie Mellon University’s Jumbotron Screen Vehicle team, as documented by Reinholtz et al. [33]. Both DARPA-winning teams were subsequently acquired by Google, and, according to Urmson et al. [34], Google’s self-driving vehicles traveled 140,000 km in California by 2010. After 2010, the development of autonomous cars accelerated rapidly, and Tesla joined the race in 2015 with its Autopilot. Despite the complexity of the process, continuous progress has produced increasingly advanced vehicles capable of understanding more road behavior. The growing importance of human–machine cooperation in transport has been highlighted in studies on the integration of autonomous and electric freight vehicles by Monios and Bergqvist [35]. Similar considerations regarding operational challenges in autonomous shipping were presented by Pijaca and Bulum [36], while Ren et al. [37] emphasized the human factor by identifying fatigue risks among professional truck drivers in evolving transport systems based on the use of one vehicle driven once by a human and once by an ego vehicle. This article is about the cooperation of a human and an autonomous vehicle.
The method results directly from the regulations on the working time of a professional driver. The driver using the time recording in the tachograph can use four variants: driving, other work, rest, and disposal. Each recorded activity has a very detailed definition, which specifies the framework and circumstances in which a driver should assign the appropriate time category. Lorencowicz et al. [38] analyzed the practical aspects of using working time in professional transport. Orzechowski et al. [39] discussed selected issues in the interpretation of driver working time regulations, while Ambrożkiewicz et al. [40] presented the theoretical and practical dimensions of truck driver working time management. For safety reasons, the regulations additionally sanction the periods of possibility to perform individual activities. In particular, “riding” can be performed only in appropriate time windows, which must be matched by appropriate variants of “rest”. The legal situation also makes it possible to create a team of drivers servicing a given vehicle. The norms and working hours of a team of drivers differ from the work of a single driver [41].
At the same time, the transport market is undergoing transformation due to the increasingly common use of vehicles with self-driving capabilities. These changes are dynamic, and the regulations are gradually adapting to the homologation and deployment of more advanced vehicles. The European Parliament has addressed this process in its 2019 resolution on autonomous driving in European transport [42]. The European Commission has also formulated a comprehensive strategy for automated mobility [43], and further legislative steps regarding driving licenses and vehicle approval have been introduced through the 2023 directive [44]. Strategic initiatives such as the GEAR 2030 report have highlighted the competitiveness and sustainable growth aspects of the automotive industry in the EU [45]. Complementary research, including the study by Barrera et al. [46], discusses trends and challenges for autonomous and unmanned vehicle technologies, while Neumann [47] presented the national perspective on the adoption of autonomous road transport in Poland. The first vehicles with homologation for the third level of autonomy appeared [48]. Passenger cars are the first to change, while modern technologies are starting to enter more and more boldly into trucks and buses.
“Driving”, whether by a human or by a vehicle in ego mode, corresponds to the same parameters and definitions. The concept of the Dynamic Driving Task (DDT) as the complete set of operations for controlling a vehicle in a given ambient scene has been described in the context of the DARPA Urban Challenge by Reinholtz et al. [33]. Motion planning and control techniques for self-driving vehicles were extensively reviewed by Paden et al. [49], while Bender et al. [50] introduced efficient map representations supporting autonomous navigation. Haavaldsen et al. [51] presented approaches to end-to-end vehicle control in simulated urban environments, and Brekke et al. [52] discussed multimodal 3D object detection as a component of advanced perception systems for autonomous driving. A professional driver who performs transport, and, at the same time, does not physically drive the vehicle, should register this working time differently from “driving” [53]. The regulations allow him to remain at “available” or to perform “other work”.
This paper presents a situation in which the driver cooperates with the ego vehicle. In the case of SAE Level 3 autonomy (Conditional Driving Automation), the emergency driver should be able to take control in the event of an emergency or dangerous situation, even though they are not the one driving the vehicle. The driver also does not “rest” or is not at “available” while driving the ego vehicle—he performs “other work”, that is, activities related to transport, but not involving the physical driving of a vehicle. The article shows simplified “working day” models for a driver and a team of drivers using a Level 3 vehicle. The models assume that the vehicle is moving throughout the “working day”, trying to stay “moving” for the maximum period. This could correspond to a situation where a driver starts his working day with a vehicle fully ready to drive and tries to drive as much of the road as possible from the current place to the destination, following the regulations.
“Other work” corresponds to different specific provisions than “driving”. The paper identifies the potential for self-driving and also discusses cooperation with a vehicle with SAE Level 4 autonomy (High Driving Automation), for which the variant of taking control (“other work”) is not necessary. The vehicle during self-drive can react to any potential threats and, in case of uncertainty, perform maneuvers to minimize risk. In such a cooperation configuration, the ego vehicle should be treated on an equal footing with the driver while driving. The driver’s time when he is not driving, but is in the course of transport, can be recorded as an “available” or “rest” variant.

4. Problem

The modern face of European transport, especially long-distance transport, poses more and more problems for carriers to solve, such as, on the one hand, constantly changing regulations: regarding cabotage requirements, making payments for drivers’ work, adaptation to the minimum hourly rate applicable in the country of transport, as well as customs and transit amendments [41]. On the other hand, the availability of new employees as drivers is decreasing. The profession of a professional driver is becoming less and less popular due to the high costs of initial training and declining prestige. The market for professional drivers is shrinking more and more, and for many years, it has been clear that there is simply a shortage of drivers in all countries. To transport goods in cars with a GVM exceeding 3.5 tons, drivers are required to undergo appropriate training. In addition to being authorized to drive trucks, they must also participate in courses enabling the transport of goods [54]. Unlike with driver’s licenses, these courses expire. Drivers must renew their courses every five years, and information about completing new training extends their ability to practice. As part of their training, drivers are familiarized with all safety and safe transport requirements, as well as the conditions that must be met in order to legally be able to perform transport. Another aspect in which it is important that the driver has the right to transport goods is any insurance case regarding damage to goods during transport. In addition to the ability to properly care for goods during transport, another leading aspect discussed during the training is the driver’s recording of their work and rest. It is important to be able to clearly determine what the driver is doing in a given time window. The driver’s time can be divided into four basic sections: driving, other work, rest, and availability [39].
The characteristics of vehicles also change dynamically. Cars used today have much lower exhaust emissions and are equipped with more and more safety systems. There is a huge, growing market for autonomous vehicles. Both society and regulations are increasingly adapting to transport solutions based on autonomous vehicles. The United Nations in Geneva in September 2020 approved a significant amendment to the 1968 Geneva Convention. Article 34 bis allows the use of autonomous vehicles. If they technically and legally comply with local regulations, ego vehicles can move within the territory of the contract country. Two new definitions have also been added in the first paragraph: “Automated Driving System” and “Dynamic Control” [55]. In the strictest sense of the word, vehicles that are not driven by humans are allowed on the roads. In the European Union, a division of vehicle autonomy is used based on the work of SAE [56]. This is a six-step scale for transitioning responsibility for driving (Figure 1), where vehicles marked “0” are driven only by drivers, and vehicles marked “5” are driven only by an artificial intelligence system [43].
In the summer of 2021, SAE Level 3 vehicles were approved for homologation in the EU [57]. In December 2021, Mercedes-Benz was the first in Europe to receive approval for the use of Level 3 automatic steering in production cars [48]. At the same time, the results presented in the EU Road Safety Report indicate that 95% of accidents are caused by human error [58]. Therefore, technical regulations for all vehicles have made it mandatory to use modern speed control systems that detect objects on the road and register the driver’s attention. Additionally, in the case of trucks and buses, blind spot detection systems and warning systems to prevent collisions with pedestrians and cyclists are mandatory [57].
Driving a vehicle can be regarded both as a task for an autonomous system (self-driving) and as a task for a human driver. The driver must, first of all, understand the scene where he is in the appropriate time window—DDT (Dynamic Driving Task). To move, of course, perception, route planning, and simultaneous vehicle operation are necessary. Successive levels of vehicle autonomy, according to SAE, mean taking more and more control over driving. The SAE Level 3 system can conditionally provide lateral control without human control, including turning, straight-ahead, and corner tracking. Additionally, the system can control the vehicle longitudinally—that is, respond appropriately to the speed depending on the situation, including acceleration and braking. Level 3 has the appropriate OEDR (Object and Event Detection and Response), along with the vehicle control; only its ODD (Operational Design Domain) is not adapted to all variants. As the ODD develops, vehicles will be able to predict the situation better and better, which will ensure their full autonomy, that is, they will be able to handle any DDT task without the need for human intervention. The formal taxonomy and definitions of these automation levels are provided in SAE J3016 [56]. Regulatory and operational considerations for freight transport applications of autonomous driving have been discussed in detail by Flämig [59]. Fundamental aspects of motion control and visual perception relevant to autonomous vehicle operation are presented by Tepteris [60].
Vehicles performing commercial transport, for which a category C, CE, or D driving license is required, are also equipped with new technologies. Most new truck models have ACC (Automatic Cruise Control), so the truck can control longitudinal driving. The vehicles additionally monitor the lane, informing the driver when a change is detected through sound systems—thanks to this, they have lateral control. Route control systems based on ACC and data from other systems are also increasingly used, including GPS, to predict the location, additionally trying to reduce fuel consumption, e.g., Pulse and Glide functionality from Scania. Trucks have long moved beyond SAE Level 1 and naturally follow developments.
Vehicles are changing. As SAE Level 3 approved passenger cars exist, the same level of commercial vehicles (VANs) will soon be on the road. The introduction of self-driving heavy vehicles is expected next.

5. Methods

This study adopts a model-based comparative analysis approach to evaluate the cooperation between professional drivers and autonomous vehicles, particularly focusing on how such collaboration impacts driver working hours and vehicle movement efficiency. The methodological framework integrates regulatory analysis, scenario-based modeling, and work-time simulation using digital tachograph categories as defined by European Union legislation.
The legal basis for modeling driver working time was derived from current European Union regulations, primarily Regulation (EC) No 561/2006, Directive (EU) 2020/1057 (Mobility Package), and corresponding national labor codes. These sources define and constrain professional driver activities within four fundamental categories: Driving—active physical operation of the vehicle; Other work—tasks such as cargo handling, documentation, vehicle checks, or supervising automated driving; Availability—time spent accompanying another driver or waiting for loading/unloading; Rest—mandatory daily or weekly rest periods.
Additionally, the SAE International J3016™ standard [61] was used to classify the automation levels of vehicles (Levels 3 and 4), and to define the operational design domain (ODD), dynamic driving task (DDT), and fallback readiness.
Three theoretical operational scenarios were developed to model how autonomous vehicle capabilities affect the working time structure of professional drivers: Scenario A: Traditional operation by a single human driver without automation; Scenario B: Cooperation with an SAE Level 3 truck (conditional automation with human fallback); Scenario C: Cooperation with an SAE Level 4 truck (high automation, no fallback required).
Each scenario assumes that the transport task begins after a full daily rest period and continues until the driver exhausts the maximum allowable working or driving time.
Model workdays were constructed based on the definitions of activity periods and rest requirements outlined in the EU legal framework. Time budgets for each category were allocated according to legal maxima and practical scheduling constraints, such as: Daily driving limit: 9 hours (extendable to 10 hours twice per week), Maximum daily working time: 13 hours (extendable to 15 hours under reduced rest), and Minimum rest: 11 consecutive hours per 24-h period (or reduced to 9 hours under specific conditions).
For each scenario, daily activity diagrams were generated to visualize the distribution of “driving”, “other work”, “availability”, and “rest” periods. Special attention was given to the treatment of automated driving time. In Level 3, self-driving periods were recorded as “other work”, as the driver must be ready to intervene. In Level 4, self-driving periods were treated as “availability” or “rest”, depending on legal interpretation and system capabilities.
Double-crew variants were also modeled to assess the combined working time potential and distribution of responsibilities in teams of two drivers.
The modeled scenarios were evaluated using the following performance indicators: total vehicle movement time per 24-h period; total effective driving time by the human driver; time allocation across tachograph categories; compliance with EU working time and rest regulations; potential for increased transport efficiency and safety.
These indicators were used to compare the impact of different levels of automation on legal compliance, operational flexibility, and the feasibility of integrating autonomous trucks into existing professional transport frameworks.
The methodological approach of this study is guided by five key research questions formulated in Section 4, which reflect the growing need to define the practical and regulatory implications of human–machine collaboration in road transport using autonomous vehicles:
Q1.
How will driving by ego vehicle affect cooperation with a person practicing the driver’s profession?
Q2.
What will the registration of the driver’s time look like, and what will the registration time for working artificial intelligence look like?
Q3.
Should an ego truck/bus have a tachograph card?
Q4.
If a professional driver undergoes training to transport goods or people safely, should buses or trucks be fully autonomous at all?
Q5.
What benefits can conditional autonomy based on constant shuttle routes for trucks bring?
To address these questions, a model-based comparative analysis was undertaken, combining legal review, scenario simulation, and structured interpretation of professional driver workday configurations using digital tachograph data categories. The objective was to assess how cooperation between human drivers and SAE Level 3/4 autonomous vehicles can affect driving time, legal compliance, and operational efficiency.
The subsequent subsections describe the legal framework, construction of representative driving scenarios, and the evaluative metrics used to compare traditional and automated operation modes in heavy-duty road transport.

Control Handover Protocols, Driver Monitoring, and Risk Scenarios

To ensure that the assessment of human–machine collaboration reflects operational reality, the study explicitly considers control handover protocols, driver monitoring mechanisms, and risk scenarios that may trigger intervention. For SAE Level 3 operation, a structured Take-Over Request (TOR) protocol is assumed. TORs are initiated in situations such as leaving the operational design domain, entering work zones, encountering adverse weather conditions, detecting sensor or system faults, or identifying complex road events such as stationary obstacles or emergency vehicles. The driver is warned using multimodal alerts that combine visual, auditory, and haptic signals, providing approximately 10–15 s of lead time for takeover. If the driver does not acknowledge the TOR within this window, the vehicle automatically initiates a minimum-risk maneuver, consisting of controlled deceleration, hazard-light activation, and stopping in a safe area. In SAE Level 4 operation, where the vehicle is capable of autonomous fallback within its designated ODD, a TOR to the human operator is optional; the system can independently execute a minimum-risk maneuver in case of confidence loss or unexpected hazards.
Driver availability is maintained through continuous monitoring by Driver Monitoring Systems (DMS). These systems observe parameters such as gaze direction, eyelid closure, blink frequency, head position, steering wheel torque, and seat occupancy to detect drowsiness or inattention. In SAE Level 3 scenarios, autonomous driving is only permitted when the driver is alert and capable of intervening; warnings escalate to TORs if signs of distraction persist. Repeated unavailability results in a temporary lockout of self-driving functions, requiring the driver to rest before resuming automated operation. In SAE Level 4 scenarios with a human attendant, DMS signals are primarily used to authorize privileged actions, maintain auditability for liability purposes, and log the presence of a qualified supervisor, even though active fallback is not strictly required.
Operational risk scenarios were also incorporated into the model to reflect realistic interruptions to movement time. High-priority risks include adverse weather conditions that degrade sensor performance, ODD exits caused by roadworks or unexpected route changes, system or sensor failures, sudden obstacles such as cut-ins or stationary vehicles, and potential cyber or electromagnetic interference affecting perception or positioning systems. Each of these scenarios is associated with specific detection cues and mitigation actions. In Level 3 operation, mitigation primarily consists of issuing a TOR to the driver; in Level 4 operation, the vehicle initiates an autonomous minimum-risk maneuver. These events may lead to temporary speed reductions or short unscheduled stops, which were represented in the simulation as minor penalties to vehicle movement time while maintaining full compliance with EU working-time regulations.
Including explicit handover logic, driver monitoring mechanisms, and risk-based interruptions strengthens the realism of the simulation and its relevance to practical deployment. It highlights that the efficiency gains of Level 3 and Level 4 automation are achievable but remain bounded by environmental disturbances, system health, and human availability. The approach also reinforces the regulatory and operational argument for treating the autonomous system as a “virtual co-driver,” where TORs, minimum-risk maneuvers, and DMS logs provide a traceable record suitable for integration with digital tachographs. This perspective aligns with emerging EU and UNECE regulations and supports future research on real-world validation and quantitative risk modeling for autonomous truck operations.

6. Experimental (Simulation) Study

Professional drivers record the course of their working day using registration in the device—tachographs. Today, it is most often a modern digital tachograph. By reading data from the tachograph, you can easily control the course of the driver’s working time. A professional driver is obliged to have one personal tachograph card, which he must use every time at work. The driver’s time from the first use of the card in the tachograph is constantly divided into sections. These are measurable periods that are recorded in such a way that there are no gaps or interruptions in time. Even if the driver removes the card from the tachograph and goes on holiday, after re-inserting the card into the tachograph, the time from the last removal of the card to the moment of reinserting it into the device will be classified as a pause. The time spent at work, or more precisely, at work transporting goods, is controlled in advance. The driver may transport goods under predetermined procedures, which include maximum time windows for appropriate activities [40]. If the driver does not comply with the requirements, there will be consequences for both him and the carrier.
Attention should also be paid to additional regulations resulting from the labor codes of the country where the transport company comes from, which also affect the way the driver works with the vehicle. For example, an employee employed under an employment contract in Poland is obliged to take a 30-min break within the first 6 h of starting work. This means that if a driver performs any work-related activity during the first 6 h, he or she must take a 30-min rest. Of course, the issues of rest can be reconciled by taking it at the same time. The obligation to rest under the labor code and the registration of rest in transport may be completed during one break. However, if a driver performs “other work” and does not drive a vehicle at work, he or she is entitled to a 30-min rest during the first 6 h.

6.1. Experimental Setup/Simulation Environment

The study was conducted as a simulation-based experiment to evaluate professional drivers and vehicle activity under different automation levels. The legal framework is defined by EU Regulation (EC) No 561/2006, which specifies the working time categories for professional drivers: driving, other work, rest, and availability.
The simulation assumes four operational configurations:
  • Single Driver (Manual Driving)—a conventional driver without vehicle automation support.
  • Double Crew (Manual Driving)—two alternating drivers, allowing extended vehicle movement without breaching working time regulations.
  • Single Driver with SAE Level 3 Automation—the driver supervises the vehicle, which can provide conditional lateral and longitudinal control; human intervention is required in specific scenarios.
  • Double Crew with SAE Level 4 Automation—high automation is available within a defined Operational Design Domain (ODD), allowing extended vehicle movement with minimal driver intervention.
The simulation environment models each scenario over a standard working day, considering the time allocation to different tachograph categories and the impact of automation on vehicle movement.
The proposed methodology is based on a discrete time simulation model that replicates professional driver activities and vehicle movement under EU working time regulations. Each minute of the workday is classified into one of four tachograph categories: driving, other work, rest, or availability. The primary objective of the simulation is to maximize daily vehicle movement time while maintaining full compliance with EU Regulation (EC) No 561/2006. In the manual operation scenarios (Single Driver and Double Crew), vehicle movement is strictly limited by the driver’s active driving time. In the SAE Level 3 and Level 4 scenarios, the model introduces periods of autonomous driving, which are assigned as availability time for the driver. The simulation algorithm optimizes the sequence of manual and automated periods to minimize vehicle idle time and ensure that the workday is scheduled in the most efficient way. This optimization strategy allows each scenario (e.g., Single Driver L3, Double Crew L4) to represent an operationally feasible and near-optimal duty cycle for the human–machine team. Although the current study focuses on operational feasibility and regulatory compliance, the modeling framework is designed for future integration with vehicle control systems. In a practical deployment, tachograph data and on-board vehicle control signals could be used to validate the model in real time and to develop multi-criteria optimization strategies that also account for safety, energy efficiency, and economic performance.

6.2. Scenario Design

The experimental study evaluates how different levels of automation influence the classification of working time and vehicle movement. Each scenario follows these principles:
  • The driver’s day is divided into periods of driving, other work, and rest/availability, consistent with tachograph recording.
  • Autonomous driving periods in SAE L3/L4 scenarios are simulated as availability time for the driver, with the vehicle maintaining movement according to the regulatory framework.
  • Daily working time constraints: Maximum daily driving time: 9–10 h (depending on regulation); Maximum working day: 15 h with a single driver; 21 h with a double crew.
  • ODD limitations for SAE L3/4: L3: Conditional control in clear road conditions, speed ≤ 60 km/h; L4: Automated control on predefined routes within ODD.

6.3. Data Processing and Metrics

The simulation produces time allocations for each activity category and total vehicle movement. The following metrics are used for analysis: Total vehicle movement time [h/day]—duration when the vehicle is in motion. Driver active driving time [h/day]—periods classified as “driving” on tachograph. Other work and rest/availability time [h/day]—supporting and non-active periods. Operational efficiency gain [%]—relative increase in vehicle movement time compared to conventional single-driver operation.
The metrics are aggregated for all four scenarios to allow direct comparison. Graphical outputs illustrate the temporal structure of the working day for each scenario, and a summary table presents the key results for quick reference.

6.4. Results and Observations

The work of a professional driver and their related regulations is an extensive topic in which the appropriate proportion of breaks (pauses) to driving time is crucial for the safety of all. The driver is obliged to take breaks while driving. This includes daily and weekly breaks. The length of rest is always controlled, and the driver should focus primarily on rest breaks. To best illustrate the working capabilities of a single driver, the next 24 h from the end of the break are indicated. Regardless of the start time, the working time of an individual driver will be settled in subsequent daily blocks (one “working day”) (Figure 2). A “driver’s week”, regardless of the day of the week on which he or she starts work, is a maximum of 6 blocks of 24 h between two weekly breaks. The values shown in Figure 2 correspond to the maximum option. The “driving” on the chart is 10 h. By default, a driver can drive a vehicle for 9 h a day, but it is possible to extend the driving time twice by one hour per calendar week. The indicated “maximum working time” is 15 h, as it is possible to shorten the daily rest three times to 9 h between weekly rests. The standard “working time” is 13 h, and daily rest is 11 h.
In addition, there are weekly and bi-weekly restrictions on driving hours. Within two calendar weeks, the driver may drive a vehicle for a maximum of 90 h. Using the maximum available driving time in one week, i.e., (2 × 10 h) + (4 × 9 h) = 56 h, in the next week, the driver will only be able to drive for 34 h. In addition, there are provisions for weekly rests, including their shortening and taking time off for shortening them, together with daily or weekly rests. Additionally, “working time” during rest breaks looks different for drivers working at night. At the same time, the latest EU Directive [41] introduces changes in weekly breaks when returning to the base.
If two employees can drive a given vehicle at the same time in a vehicle over 3.5 tons, they are obliged to keep joint records of working time. A double crew of drivers is created to operate a given vehicle, the working time of which is shown in Figure 3. The values shown in Figure 3 also represent the maximum variant. The main difference between the work of a team of drivers and the work of a single driver is the extension of the “working day” to 30 h. At the same time, the “maximum working time” is extended to 21 h. Drivers can take turns driving the vehicle, and the driver sitting in the passenger seat can take a “break” if he or she is not performing “other work” at that time. Otherwise, they maintain an “available” status.
The current situation in the regulations means that both drivers and carriers must be fluent in understanding the regulations on working time records. Regulations based on calendar days and weeks, as well as regulations based on “working days”—or rather, times between daily or weekly rest periods—do not facilitate efficient navigation of the regulations.
The regulations focus mainly on two aspects: driving and resting, defining them very clearly. According to the regulations, driving is the time spent on steering a vehicle. To define it even more precisely, this is the time window in which all physical activities related to driving a vehicle are performed. The driver uses perception to control the moving vehicle longitudinally and laterally in a given time window. The regulations also define “other work”, i.e., any other type of activity related to transport or transporting other than driving a vehicle. This includes all activities related to cargo securing and fulfillment of documentation obligations. For example, the driver registers “other work” during loading, customs clearance, or road control, also during the preparation of the vehicle for the road, e.g., refueling or cleaning the cargo area. Practical aspects of professional driver time allocation and its proper registration were analyzed by Lorencowicz et al. [38], while Kusminska-Fijalkowska et al. [62] discussed related safety and operational considerations in the context of transport activities.
The task of autonomous driving (DDT) is the same as driving included in the regulations for professional drivers. Only the type of ODD environment recognized by OEDR determines the degree of autonomy. The actual physical driving of the autonomous vehicle, i.e., the decision on subsequent movements, is made by the controller, who receives information from the motion planning module. The motion planning module receives data from the environmental perception module and the mapping module, and these receive a set of raw measurements from the exteroceptive and proprioceptive sensors. This classic version of software architecture also includes a supervisory unit that checks the operation of hardware, software, and sensors to minimize errors and ensure the most comfortable driving maneuvers.
If a driver’s “driving” and an ego vehicle’s “driving” are equal, should a professional driver record the time a truck drives itself as “driving”?
Figure 4 and Figure 5 show the work schedule of a driver driving a truck with SAE Level III autonomy. This level, “Conditional Driving Automation”, requires the driver to take control in the event of a failure. At the same time, it does not require the driver to constantly focus on the road. At SAE Level 3, ODD is limited, and a driver is needed to take over if an unforeseen event is detected. Self-driving time is presented as “other work”. In Figure 4, attempts were made to maximize the “driving” available to the driver. In Figure 5, maximum use was made of the time available for self-driving (“other work”).
The “movement time” (13.25 h) shown in Figure 4 consists of the maximum driving time (10 h) and “other work” (3.25 h). Figure 5 shows a variant where the driver only starts and finishes the trip before the end of the maximum working time (2 × 30 min of working time). In this variant, the “movement time” (12.75 h) consists of “driving” (1 h) and “other work” (11.75 h). The maximum available operating time for both variants is 15 h.
Figure 6 also shows the maximum values using an ego Level 3 vehicle but with a double crew on the vehicle. Here, drivers perform “other work” alternately with the “availability” phase, while maintaining mandatory breaks under the labor code. While the “movement time” with a double crew increased by only 1 h, the use of the ego vehicle significantly reduced the “driving” recorded by the drivers.
Table 1 summarizes the key parameters of all modeled scenarios, including total vehicle movement time, driver’s active driving, and periods classified as other work or rest. This comparative overview highlights the potential operational benefits of integrating SAE Level 3 and 4 automation into professional transport.
In vehicles with SAE Level 4 approval (High Automation Driving), the system can respond to all ODD variants, achieving the lowest possible risk, even without driver participation, in an emergency. This means that the vehicle can be steered alternately, and if necessary, a professional driver can take over, but the system will respond to emergencies, avoiding risks (usually by stopping on the side of the road). The course of cooperation between such a road set and a professional driver will best suit a situation in which there are currently two drivers in the set (Figure 3). In the case of Level 5 autonomous vehicles, cooperation will not occur because the vehicle has unlimited ODD. This means that the journey takes place completely without the participation of a professional driver, and the system is able to respond correctly to any situation encountered on the road. The professional driver’s other responsibilities, including ensuring the safety of the goods, are controversial. The introduction of Level 5 vehicles for use in transport must certainly be preceded by statutory changes and social transformation.

7. Discussion

The comparative analysis conducted in this study demonstrates that collaboration between professional drivers and autonomous vehicles—particularly those at SAE Levels 3 and 4—can significantly influence both the effective movement time of the vehicle and the formal classification of the driver’s working hours. The simulation scenarios provide insights into how self-driving functionality can be operationally integrated within the existing regulatory framework, while also raising normative and technical challenges. One of the most prominent findings is that in vehicles operating under SAE Level 3, where conditional automation requires driver fallback readiness, periods of automated driving can be reasonably registered as “other work”. This complies with EU Regulation (EC) No 561/2006 and supports the interpretation that supervising autonomous operation constitutes a task within the scope of transport activity. For SAE Level 4, where no human intervention is required within the designated ODD, the vehicle can be regarded as functionally independent. In such cases, time allocation in tachograph records becomes ambiguous. While current regulations do not provide explicit mechanisms for assigning a digital tachograph card to an autonomous system, the concept of “team driving” suggests a legal path forward. The idea of treating the ego vehicle as a “virtual co-driver” opens the possibility of distributing working time between a human and a machine, though this requires legislative updates or technical adaptations in tachograph systems. This ambiguity leads to a broader question (Q3): Should ego vehicles possess their own digital identification within tachograph systems? From a legal and auditing perspective, this would enhance traceability and align with existing data retention protocols. However, the deployment of such a system will necessitate international harmonization and likely changes to the AETR convention.
The simulation results show that the role of the professional driver remains highly relevant, especially in the transitional phase before Level 5 systems become legally and operationally viable. While full autonomy eliminates the need for human intervention in driving, it does not remove the broader responsibilities tied to professional transport: cargo integrity, passenger safety, documentation, and compliance verification. Thus, as addressed in Q4, mandatory training for professional drivers continues to be essential—not necessarily for operating the vehicle, but for managing the transport process in hybrid scenarios. This is particularly important in emergency situations where a human must take over, or in jurisdictions where liability and insurance systems are yet to adapt to full autonomy.
Scenarios involving fixed shuttle routes suggest significant benefits from Level 4 autonomy, including: Extended movement time without violating driver work-time restrictions, reduced exposure to driver fatigue and human error, and smoother integration with transport schedules and logistics networks. In practical terms, self-driving periods effectively “unlock” additional movement time. For SAE Level 3, the model shows a 3.25-h gain, while Level 4 autonomy increases vehicle activity by up to 14.25 h in double-crew configurations, without exceeding labor code limitations. This gain could allow for longer distances or slower, fuel-optimized driving, supporting decarbonization goals.
Nevertheless, these theoretical advantages depend on well-defined ODDs and reliable fallback protocols. Without clear boundaries for autonomous operation, the practical deployment of ego vehicles will face major safety and legal concerns. Moreover, public perception and driver acceptance remain crucial, as highlighted in prior studies.
The safe deployment of semi-autonomous heavy vehicles requires not only compliance with regulatory frameworks but also a clear understanding of operational risk. The safety of the proposed method and existing solutions depends on the interaction of technical, environmental, and human factors. Based on the current literature and operational considerations, four primary risk scenarios have been identified: (1) sensor failure or misperception due to adverse weather, (2) electromagnetic interference (EMI) or intentional electromagnetic attacks affecting power electronics or visual recognition, (3) cyberattacks such as GPS spoofing or denial-of-service (DoS), and (4) delayed driver reaction in SAE Level 3 operation.
Each scenario has been qualitatively assessed in terms of event severity and probability of occurrence, as shown in Table 2. Severity reflects the potential impact on vehicle safety and cargo integrity, while probability reflects the likelihood of occurrence under current technology and operational practices.
This qualitative analysis indicates that the highest-risk events are associated with environmental perception errors and delayed driver reactions. While the probability of intentional electromagnetic or cyberattacks remains relatively low, their severity justifies the adoption of human-aided strategies, real-time monitoring, and redundant fail-safe protocols. This approach aligns with the systematic three-stage safety enhancement framework applied in unmanned aerial vehicle motor drive and gimbal systems, which emphasizes hazard identification, event likelihood estimation, and mitigation planning.
Future work will focus on extending this qualitative assessment into a quantitative risk analysis, integrating failure probabilities from real-world sensor data, communication reliability statistics, and human reaction times to provide a comprehensive safety validation framework for autonomous truck operations.

8. Conclusions

This study investigated the operational feasibility of human–machine collaboration in professional road transport using SAE Level 3 and Level 4 autonomous trucks under the framework of EU working time regulations. The simulation-based analysis provided insights into vehicle movement time, driver workload, and the potential operational benefits of integrating automation with human supervision.
The research questions [Q1–Q5] were addressed as follows:
[Q1]
How will ego vehicle driving affect co-operation with a person practicing the driver’s profession? If the vehicle transports goods or persons, provisions must be made regarding liability for potential damage. Driving time in an autonomous system should be recorded in the same way as professional driving time. Time recording can be carried out using current methods, i.e., digital tachographs.
[Q2]
What will the registration of the driver’s time look like, and what will the registration time for working artificial intelligence look like?
[Q3]
Should an ego truck/bus have a tachograph card? Depending on the degree of vehicle autonomy, the ego driving time should be recorded in the tachograph using other permissible variants. In the case of vehicles with SAE Level 3 autonomy, this registration should be done using “other work”, as the vehicle remains, to some extent, under the control of the professional driver. For SAE Level 4, the driving record should be identical to the mode of recording the working time of a professional driver, i.e., recorded as “driving”, assigned to the ego vehicle. However, there are two fundamental unresolved questions here: Does the driving of an SAE Level 4 ego vehicle require replacement by a professional driver? Can such a vehicle be driven by a double crew of two professional drivers? A partial answer to these questions is provided by the regulations regarding the registration of professional drivers’ work. The resulting possibilities of working time, including the provisions of the Labor Code, for which breaks are required, and the cooperation of the ego vehicle should be adapted to the professional driver and their needs for breaks. However, vehicles that are characterized by a higher degree of automation and, at the same time, those that will professionally transport goods or people should record working time on an equal footing with the driver. Due to the possibility of building a double crew of two professional drivers, trucks must be able to completely disable the ego function. Moreover, drivers who will cooperate with the ego vehicle should additionally undergo mandatory training, during which the variants of responsibilities and duties resulting from cooperation with an ego vehicle will be presented. This includes, in particular, the securing of goods or passengers.
[Q4]
If a professional driver undergoes training to transport goods or people safely, should buses or trucks be fully autonomous at all? Yes, today’s autonomous cars are typically vehicles designed for commercial passenger transportation. This technology is used by various taxi service companies or their variants. Since commercial transport by ego passenger cars has been allowed, it seems a natural consequence that trucks and buses will also be fully autonomous in the future. The biggest barrier to the use of heavy autonomous vehicles seems to be the potential consequences of possible errors. Here, at least until the times of social transformation, cooperation between machines and humans could reduce social unrest and get society used to the sight of heavy vehicles independently carrying out the task of driving.
[Q5]
What benefits can conditional autonomy based on constant shuttle routes for trucks bring? There are many advantages. Autonomous vehicles make decisions while driving, without causing boredom or distraction. They are created in accordance with the ego vehicle safety standards certified by ISO. Additionally, the NHTSA guidelines, which are followed by the creators of autonomous vehicles, ensure that the most common sources of threats are reduced to the lowest possible risk level. The first benefit is therefore increased safety in transport. When an SAE Level 3 truck is self-driving, the driver is performing “other work,” according to the models. In the case of SAE Level 4, the cooperation between a vehicle and a professional driver is similar to today’s double crew. This means that in each variant, the relative movement time of the vehicle increases. Combining the driver’s and the ego vehicle’s potential driving times results in a longer total movement time. For an SAE Level 3 vehicle, this time is extended to a maximum value of 13:25, while for an SAE Level 4 vehicle, it is increased by an hour. With a longer driving time, two scenarios are possible: either you can transport a longer distance, or you can transport a similar distance but at a slower speed, which, of course, will directly affect the ecological index. There are also problems in transport companies with the implementation of transport in key fixed directions. Where transport is difficult due to distance and weather conditions, the introduction of ego vehicles in transport companies would reduce the potential risk of a lack of deliveries in these directions. The fact that transport takes place on fixed shuttle routes means that building an appropriate ODD should be easier. Therefore, for transport performed along fixed routes, e.g., between terminals, the use of vehicles with higher automation will significantly affect the possibilities of providing transport services.
Limitations and Future Challenges: This study is based on simulation only and has not been validated in real-world transport operations. The analysis focused solely on SAE Level 3 and 4, excluding full automation (Level 5). Quantitative evaluation of safety, energy efficiency, and economic impact remains outside the scope of this initial study. Future research should focus on experimental validation, risk quantification, and integration of autonomous system logs with digital tachographs and driver monitoring systems.
The findings provide a structured framework for the gradual adoption of autonomous trucks in professional road transport and underline the operational value of human–machine collaboration. Addressing the identified limitations through real-world trials and quantitative performance evaluation will form the basis of future research.
The presented analysis highlights the importance of aligning autonomous vehicle operation with driver work regulations, tachograph data recording, and hybrid human–machine operational models. The study also emphasizes the regulatory ambiguity related to classifying self-driving periods, suggesting that the introduction of a “virtual co-driver” concept or digital tachograph identification for autonomous systems may be required in the future. Future research should focus on experimental validation of the proposed working time and movement models in real transport operations; development of tachograph standards that integrate autonomous system activity logs; assessment of the safety, energy efficiency, and labor market impact of large-scale SAE Level 4 deployment; exploration of the transition toward SAE Level 5 autonomy and its implications for regulatory frameworks and logistics networks. The findings provide a quantitative foundation for transport operators, regulators, and vehicle manufacturers to plan the phased adoption of autonomous trucks and to optimize human–machine collaboration in professional road transport.

Author Contributions

Conceptualization, T.N. and R.Ł.; Methodology, T.N.; Formal analysis, T.N.; Data curation, R.Ł.; Writing—original draft, R.Ł.; Writing—review & editing, T.N.; Funding acquisition, T.N. All authors have read and agreed to the published version of the manuscript.

Funding

This study was funded by the Gdynia Maritime University, under the research project: WN/2025/PZ/07.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Fu, H.; Fu, X.; Ye, S.; Wang, K.; Chen, T. Understanding Preferences for Autonomous Trucks Functions in China: Insights from Drivers and Organizational Buyers. Transp. Policy 2025, 170, 110–119. [Google Scholar] [CrossRef]
  2. Bhoopalam, A.K.; van den Berg, R.; Agatz, N.; Chorus, C.G. The Long Road to Automated Trucking: Insights from Driver Focus Groups. Transp. Res. Part C Emerg. Technol. 2023, 156, 104351. [Google Scholar] [CrossRef]
  3. Wang, M.; Mehrotra, S.; Wong, N.; Parker, J.; Roberts, S.C.; Kim, W.; Romo, A.; Horrey, W.J. Human-Machine Interfaces and Vehicle Automation: A Review of the Literature and Recommendations for System Design, Feedback, and Alerts. Transp. Res. Part F Traffic Psychol. Behav. 2024, 107, 549–561. [Google Scholar] [CrossRef]
  4. Sekadakis, M.; Katrakazas, C.; Clement, P.; Prueggler, A.; Yannis, G. Safety and Impact Assessment for Seamless Interactions through Human-Machine Interfaces: Indicators and Practical Considerations. Transp. Res. Procedia 2023, 72, 2425–2432. [Google Scholar] [CrossRef]
  5. Zhao, Z.; Liu, T.; Zhang, L.; Xie, S.; Jin, H. Enhancing Autonomous Vehicle Safety: An Integrated Ensemble Learning-Logit Model for Accident Severity Prediction and Analysis. Int. J. Transp. Sci. Technol. 2025. [Google Scholar] [CrossRef]
  6. Wang, J.; Pant, Y.V.; Jiang, Z. Learning-Based Modeling of Human-Autonomous Vehicle Interaction for Improved Safety in Mixed-Vehicle Platooning Control. Transp. Res. Part C Emerg. Technol. 2024, 162, 104600. [Google Scholar] [CrossRef]
  7. Castritius, S.-M.; Lu, X.-Y.; Bernhard, C.; Liebherr, M.; Schubert, P.; Hecht, H. Public Acceptance of Semi-Automated Truck Platoon Driving. A Comparison between Germany and California. Transp. Res. Part F Traffic Psychol. Behav. 2020, 74, 361–374. [Google Scholar] [CrossRef]
  8. Kabbani, T.; Kim, S.; Serbes, D.; Ozan, B.; van Egmond, R.; Hartavi, A.E. Improved Autonomous Trucker-Vehicle Dialogue under Critical Scenarios through Fluid-HMI. Transp. Res. Procedia 2023, 72, 674–680. [Google Scholar] [CrossRef]
  9. Gwak, J.; Shimono, K.; Suda, Y. How Encounter Timing Affects the Impact of eHMI on Surrounding Drivers during Automated Truck Merging. Transp. Res. Part F Traffic Psychol. Behav. 2025, 110, 182–194. [Google Scholar] [CrossRef]
  10. Brill, S.; Payre, W.; Debnath, A.; Horan, B.; Birrell, S. External Human–Machine Interfaces for Automated Vehicles in Shared Spaces: A Review of the Human–Computer Interaction Literature. Sensors 2023, 23, 4454. [Google Scholar] [CrossRef] [PubMed]
  11. Fonseca, T.; Ferreira, S. Monitoring Technologies for Truck Drivers: A Systematic Review of Safety and Driving Behavior. Appl. Sci. 2025, 15, 6513. [Google Scholar] [CrossRef]
  12. Yan, S.; Wei, Q.; Xie, X.; Zhao, D.; Liu, X. Driver–Automated Cooperation Driving Authority Optimization Framework for Shared Steering Control. Processes 2024, 12, 2313. [Google Scholar] [CrossRef]
  13. Kim, E.; Kim, Y.; Park, J. The Necessity of Introducing Autonomous Trucks in Logistics 4.0. Sustainability 2022, 14, 3978. [Google Scholar] [CrossRef]
  14. Kim, H.; Ko, J.; Oh, C.; Kim, S. Evaluation of Autonomous Driving Safety by Operational Design Domains (ODD) in Mixed Traffic. Sustainability 2024, 16, 9672. [Google Scholar] [CrossRef]
  15. Guo, J.; Yuan, Q.; Yu, J.; Chen, X.; Yu, W.; Cheng, Q.; Wang, W.; Luo, W.; Jiang, X. External Human–Machine Interfaces for Autonomous Vehicles from Pedestrians’ Perspective: A Survey Study. Sensors 2022, 22, 3339. [Google Scholar] [CrossRef] [PubMed]
  16. Alhawiti, A.; Kwigizile, V.; Oh, J.-S.; Asher, Z.D.; Hakimi, O.; Aljohani, S.; Ayantayo, S. The Effectiveness of eHMI Displays on Pedestrian–Autonomous Vehicle Interaction in Mixed-Traffic Environments. Sensors 2024, 24, 5018. [Google Scholar] [CrossRef]
  17. Agredo-Delgado, V.; Ruiz, P.H.; Collazos, C.A.; Moreira, F. An Exploratory Study on the Validation of THUNDERS: A Process to Achieve Shared Understanding in Problem-Solving Activities. Informatics 2022, 9, 39. [Google Scholar] [CrossRef]
  18. Zheng, Y.; Ren, X. Developing a Multimodal HMI Design Framework for Automotive Wellness in Autonomous Vehicles. Multimodal Technol. Interact. 2022, 6, 84. [Google Scholar] [CrossRef]
  19. Visconti, P.; Rausa, G.; Del-Valle-Soto, C.; Velázquez, R.; Cafagna, D.; De Fazio, R. Innovative Driver Monitoring Systems and On-Board-Vehicle Devices in a Smart-Road Scenario Based on the Internet of Vehicle Paradigm: A Literature and Commercial Solutions Overview. Sensors 2025, 25, 562. [Google Scholar] [CrossRef] [PubMed]
  20. Zhang, Y.; Carballo, A.; Yang, H.; Takeda, K. Perception and Sensing for Autonomous Vehicles under Adverse Weather Conditions: A Survey. ISPRS J. Photogramm. Remote Sens. 2023, 196, 146–177. [Google Scholar] [CrossRef]
  21. Kim, J.; Park, B.; Kim, J. Empirical Analysis of Autonomous Vehicle’s LiDAR Detection Performance Degradation for Actual Road Driving in Rain and Fog. Sensors 2023, 23, 2972. [Google Scholar] [CrossRef] [PubMed]
  22. Jie, H.; Zhao, Z.; Zeng, Y.; Chang, Y.; Fan, F.; Wang, C.; See, K.Y. A Review of Intentional Electromagnetic Interference in Power Electronics: Conducted and Radiated Susceptibility. IET Power Electron. 2024, 17, 1487–1506. [Google Scholar] [CrossRef]
  23. Kim, K.; Kim, J.S.; Jeong, S.; Park, J.-H.; Kim, H.K. Cybersecurity for Autonomous Vehicles: Review of Attacks and Defense. Comput. Secur. 2021, 103, 102150. [Google Scholar] [CrossRef]
  24. Petit, J.; Shladover, S.E. Potential Cyberattacks on Automated Vehicles. IEEE Trans. Intell. Transp. Syst. 2015, 16, 546–556. [Google Scholar] [CrossRef]
  25. Liu, T.; Liu, F. Graph Neural Networks for Aerodynamic Analysis of Truck Platoon Under Heterogeneous Traffic Environment. IEEE Access 2025, 13, 112978–112985. [Google Scholar] [CrossRef]
  26. Liu, T.; Meidani, H. Neural Network Surrogate Models for Aerodynamic Analysis in Truck Platoons: Implications on Autonomous Freight Delivery. Int. J. Transp. Sci. Technol. 2024, 16, 234–243. [Google Scholar] [CrossRef]
  27. Dickmanns, E.D.; Zapp, A. A Curvature-Based Scheme for Improving Road Vehicle Guidance by Computer Vision. In Mobile Robots; SPIE: Washington, WA, USA, 1987; Volume 0727, pp. 161–168. [Google Scholar]
  28. Dickmanns, E.D. The Development of Machine Vision for Road Vehicles in the Last Decade. In Proceedings of the Intelligent Vehicle Symposium, Versailles, France, 17–21 June 2002; IEEE: New York, NY, USA, 2003; Volume 1, pp. 268–281. [Google Scholar]
  29. Thorpe, C.; Herbert, M.; Kanade, T.; Shafer, S. Toward Autonomous Driving: The CMU Navlab. I. Perception. IEEE Expert 1991, 6, 31–42. [Google Scholar] [CrossRef]
  30. Tsugawa, S.; Jeschke, S.; Shladover, S.E. A Review of Truck Platooning Projects for Energy Savings. IEEE Trans. Intell. Veh. 2016, 1, 68–77. [Google Scholar] [CrossRef]
  31. Kavathekar, P.; Chen, Y. Vehicle Platooning: A Brief Survey and Categorization. In IDETC-CIE2011 ASME/IEEE International Conference on Mechatronic and Embedded Systems and Applications, Parts A and B, Washington, DC, USA, 28 August 2011; American Society of Mechanical Engineers: New York, NY, USA, 2012; Volume 3, pp. 829–845. [Google Scholar]
  32. Wang, Q.; Chen, H.; Gong, J.; Zhao, X.; Li, Z. Studying Driver’s Perception Arousal and Takeover Performance in Autonomous Driving. Sustainability 2023, 15, 445. [Google Scholar] [CrossRef]
  33. Reinholtz, C.; Alberi, T.; Anderson, D.; Bacha, A.; Bauman, C.; Cacciola, S.; Currier, P.; Dalton, A.; Farmer, J.; Faruque, R.; et al. DARPA Urban Challenge Technical Paper; University of Sydney: Sydney, Australia, 2007. [Google Scholar]
  34. Urmson, C.; Anhalt, J.; Bagnell, D.; Baker, C.; Bittner, R.; Clark, M.N.; Dolan, J.; Duggins, D.; Galatali, T.; Geyer, C.; et al. Autonomous Driving in Urban Environments: Boss and the Urban Challenge. J. Field Robot. 2008, 25, 425–466. [Google Scholar] [CrossRef]
  35. Monios, J.; Bergqvist, R. The Transport Geography of Electric and Autonomous Vehicles in Road Freight Networks. J. Transp. Geogr. 2019, 80, 102500. [Google Scholar] [CrossRef]
  36. Pijaca, M.; Bulum, B. Comparison of Problems Related to the Carriage of Goods by Sea Between Traditional and Autonomous Vessels. TransNav Int. J. Mar. Navig. Saf. Sea Transp. 2021, 15, 125–131. [Google Scholar] [CrossRef]
  37. Ren, X.; Pritchard, E.; van Vreden, C.; Newnam, S.; Iles, R.; Xia, T. Factors Associated with Fatigued Driving among Australian Truck Drivers: A Cross-Sectional Study. Int. J. Environ. Res. Public Health 2023, 20, 2732. [Google Scholar] [CrossRef] [PubMed]
  38. Lorencowicz, E.; Jarmuł, R.; Koszel, M.; Przywara, A. Analiza wykorzystania czasu pracy kierowców. Probl. Transp. I Logistyki 2017, 38, 27–38. [Google Scholar] [CrossRef]
  39. Orzechowski, M.; Mrozik, M.; Danilecki, K. Selected Aspects of Analysis of Driver?S Working Time. AUTOBUSY–Tech. Eksploat. Syst. Transp. 2018, 19, 205–207. [Google Scholar] [CrossRef]
  40. Ambrożkiewicz, B.; Przystupa, K.; Nawłatyna, B. Czas pracy kierowcy samochodu ciężarowego: Teoria, a praktyka. Autobusy Tech. Eksploat. Syst. Transp. 2018, 19, 855–858. [Google Scholar] [CrossRef]
  41. Regulation (EU) 2020/1054 of The European Parliament and of The Council of 15 July 2020. Official Journal of the European Union 2020, L 249/17 2020. Available online: https://eur-lex.europa.eu/eli/reg/2020/1054/oj/eng (accessed on 1 July 2025).
  42. Autonomous Driving in European Transport, European Parliament Resolution of 15 January 2019 on Autonomous Driving in European Transport (2018/2089(INI)). Official Journal of the European Union 2020, C 411/2 2020. Available online: https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52019IP0005 (accessed on 1 July 2025).
  43. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee, the Committee of the Regions on the Road to Automated Mobility: An EU Strategy for Mobility of the Future 2018. Available online: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:52018DC0283 (accessed on 1 July 2025).
  44. Directive of the European Parliament and of the Council on Driving Licences, Amending Directive (EU) 2022/2561 of the European Parliament and of the Council Regulation (EU) 2018/1724 of the European Parliament and of the Council and Repealing Directive 2006/126/EC of the European Parliament and of the Council and Commission Regulation (EU) No 383/2012 2023. Available online: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:52023PC0127 (accessed on 1 July 2025).
  45. GEAR 2030 High Level Group on the Competitiveness and Sustainable Growth of the Automotive Industry in the European Union 2017. Available online: https://www.europarl.europa.eu/cmsdata/141562/GEAR%202030%20Final%20Report.pdf (accessed on 1 July 2025).
  46. Barrera, C.; Padron Armas, I.; Luis, F.; Llinas, O.; Marichal, N. Trends and Challenges in Unmanned Surface Vehicles (USV): From Survey to Shipping. TransNav Int. J. Mar. Navig. Saf. Sea Transp. 2021, 15, 135–142. [Google Scholar] [CrossRef]
  47. Neumann, T. Perspektywy wykorzystania pojazdów autonomicznych w transporcie drogowym w Polsce. Autobusy Tech. Eksploat. Syst. Transp. 2018, 19, 787–794. [Google Scholar] [CrossRef]
  48. Mercedes-Benz Receives World’s First Internationally Valid System Approval for Conditionally Automated Driving. Available online: http://media.mbusa.com/releases/mercedes-benz-receives-worlds-first-internationally-valid-system-approval-for-conditionally-automated-driving (accessed on 20 August 2023).
  49. Paden, B.; Cap, M.; Yong, S.Z.; Yershov, D.; Frazzoli, E. A Survey of Motion Planning and Control Techniques for Self-Driving Urban Vehicles. IEEE Trans. Intell. Veh. 2016, 1, 33–55. [Google Scholar] [CrossRef]
  50. Bender, P.; Ziegler, J.; Stiller, C. Lanelets: Efficient Map Representation for Autonomous Driving. In Proceedings of the 2014 IEEE Intelligent Vehicles Symposium Proceedings, Dearborn, MI, USA, 8–11 June 2014; pp. 420–425. [Google Scholar]
  51. Haavaldsen, H.; Aasbø, M.; Lindseth, F. Autonomous Vehicle Control: End-to-End Learning in Simulated Urban Environments. In Nordic Artificial Intelligence Research and Development; Bach, K., Ruocco, M., Eds.; Springer International Publishing: Cham, Switzerland, 2019; pp. 40–51. [Google Scholar]
  52. Brekke, Å.; Vatsendvik, F.; Lindseth, F. Multimodal 3D Object Detection from Simulated Pretraining. In Nordic Artificial Intelligence Research and Development; Bach, K., Ruocco, M., Eds.; Springer International Publishing: Cham, Switzerland, 2019; pp. 102–113. [Google Scholar]
  53. Zaranka, J.; Pečeliūnas, R.; Žuraulis, V. A Road Safety-Based Selection Methodology for Professional Drivers: Behaviour and Accident Rate Analysis. Int. J. Environ. Res. Public Health 2021, 18, 12487. [Google Scholar] [CrossRef]
  54. Wolak, M. Problem braku kierowców zawodowych. Rekomendowane rozwiązania. In Polski Instytut Transportu Drogowego; Polski Instytut Transportu Drogowego: Wrocław, Polska, 2018. [Google Scholar]
  55. Economic Commission for Europe Inland Transport Committee Global Forum for Road Traffic Safety Report of the Global Forum for Road Traffic Safety on Its Eighty-First Session. United Nations ECE/TRANS/WP.1/173/Add.1 2020. Available online: https://unece.org/sites/default/files/2021-01/ECE-TRANS-WP.1-173-Add1e.pdf (accessed on 1 July 2025).
  56. Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles. SAE International Surface Vehicle Recommended Practice 2021, J3016TM 2021. Available online: https://www.sae.org/standards/content/j3016_202104/ (accessed on 1 July 2025).
  57. Regulation (EU) 2019/2144 of the European Parliament and of the Council of 27 November 2019. Official Journal of the European Union 2019, L 325/1 2019. Available online: https://eur-lex.europa.eu/eli/reg/2019/2144/oj/eng (accessed on 1 July 2025).
  58. Report from the Commission to the European Parliament and the Council Saving Lives: Boosting Car Safety in the EU 2016. Available online: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52016DC0787 (accessed on 1 July 2025).
  59. Flämig, H. Autonomous Vehicles and Autonomous Driving in Freight Transport. In Autonomous Driving Technical Legal and Social Aspects; Maurer, M., Gerdes, J.C., Lenz, B., Winner, H., Eds.; Springer: Berlin/Heidelberg, Germany, 2016; pp. 365–385. ISBN 978-3-662-48847-8. [Google Scholar]
  60. Tepteris, G.V. Autonomous Vehicles: Basic Concepts in Motion Control and Visual Perception. Master’s Thesis, University of the Aegean School of Engineering, Chios, Greece, 2020. [Google Scholar]
  61. SAE Standard J3016; Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems. SAE International: Warrendale, PA, USA, 2021.
  62. Kusminska-Fijalkowska, A.; Lukasik, Z.; Kozyra, J.; Olszanska, S. Diagnostics of Causes, Needs and Options of Improvement of Security in Railway Transport. TransNav Int. J. Mar. Navig. Saf. Sea Transp. 2022, 16, 343–351. [Google Scholar] [CrossRef]
Figure 1. Levels of automation. Source: European Commission, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee, the Committee of the Regions, and On the road to automated mobility: An EU strategy for mobility of the future, p. 3.
Figure 1. Levels of automation. Source: European Commission, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee, the Committee of the Regions, and On the road to automated mobility: An EU strategy for mobility of the future, p. 3.
Energies 18 04219 g001
Figure 2. Maximum work time of the driver.
Figure 2. Maximum work time of the driver.
Energies 18 04219 g002
Figure 3. Maximum working time of drivers in a double crew.
Figure 3. Maximum working time of drivers in a double crew.
Energies 18 04219 g003
Figure 4. Maximum work time of the driver and the SAE Level 3 ego truck—“driving” mode.
Figure 4. Maximum work time of the driver and the SAE Level 3 ego truck—“driving” mode.
Energies 18 04219 g004
Figure 5. Maximum work time of the driver and the SAE Level 3 ego truck—“other work” mode.
Figure 5. Maximum work time of the driver and the SAE Level 3 ego truck—“other work” mode.
Energies 18 04219 g005
Figure 6. Maximum work time of drivers in a double crew and an SAE Level 3 ego truck.
Figure 6. Maximum work time of drivers in a double crew and an SAE Level 3 ego truck.
Energies 18 04219 g006
Table 1. Comparison of vehicle movement and driver activity time across all modeled scenarios.
Table 1. Comparison of vehicle movement and driver activity time across all modeled scenarios.
ScenarioVehicle Movement
[h]
Driving
[h]
Other Work
[h]
Rest/Availability
[h]
Single driver (no AV)10.0010.003.0011.00
Single driver + L313.2510.003.2510.75
Single driver + L413.751.000.0013.00
Double crew + L314.2510.004.259.75
Double crew + L414.251.000.0014.00
Table 2. Identified safety risk scenarios with event severity, probability, and mitigation strategies.
Table 2. Identified safety risk scenarios with event severity, probability, and mitigation strategies.
Risk ScenarioSeverityProbabilityMitigation Strategy
Sensor failure in fog/rain/snowHighMediumHuman intervention (SAE L3), sensor redundancy
Electromagnetic interference (EMI)MediumLowShielding, fail-safe stop protocols
Cyberattack (GPS spoofing, DoS)HighLowSystem monitoring, AI fallback, black box logging
Delayed driver reaction in SAE Level 3HighMediumDriver training, attention monitoring systems
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Neumann, T.; Łukasik, R. Integrating Autonomous Trucks into Human-Centric Operations: A Path to Safer and More Energy-Efficient Road Transport. Energies 2025, 18, 4219. https://doi.org/10.3390/en18164219

AMA Style

Neumann T, Łukasik R. Integrating Autonomous Trucks into Human-Centric Operations: A Path to Safer and More Energy-Efficient Road Transport. Energies. 2025; 18(16):4219. https://doi.org/10.3390/en18164219

Chicago/Turabian Style

Neumann, Tomasz, and Radosław Łukasik. 2025. "Integrating Autonomous Trucks into Human-Centric Operations: A Path to Safer and More Energy-Efficient Road Transport" Energies 18, no. 16: 4219. https://doi.org/10.3390/en18164219

APA Style

Neumann, T., & Łukasik, R. (2025). Integrating Autonomous Trucks into Human-Centric Operations: A Path to Safer and More Energy-Efficient Road Transport. Energies, 18(16), 4219. https://doi.org/10.3390/en18164219

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop