Dynamic Integration of Collaborative Robots and Automated Guided Vehicles In Manufacturing Systems – A Simulation Study

Kuppan Chetty Ramanathan , Koppaka Sailendra, Harsha Kumara Pappula, Tharun Thirumalai and Raja Narayanan
School of Mechanical Engineering SASTR, Thanjavur, Tamil Nadu, India
Correspondence to: Kuppan Chetty Ramanathan, kuppanchetty@mech.sastra.edu

DOI: https://doi.org/10.70389/PJS.100136

Additional information

• Ethical approval: N/a
• Consent: N/a
• Funding: No industry funding
• Conflicts of interest: N/a
• Author contribution: Kuppan Chetty Ramanathan, Koppaka Sailendra, Harsha Kumara Pappula, Tharun Thirumalai and Raja Narayanan – Conceptualization, Writing – original draft, review and editing
• Guarantor: Kuppan Chetty Ramanathan
• Provenance and peer-review:
  Unsolicited and externally peer-reviewed
• Data availability statement: N/a

Keywords: agv-mounted cobots, Changeable magnetic end effector, Coppeliasim manufacturing simulation, Diesel engine assembly automation, Line-following navigation algorithm.

Peer Review
Received: 13 August 2025
Last revised: 25 September 2025
Accepted: 29 September 2025
Version accepted: 3
Published: 16 October 2025

Plain Language Summary Infographic

Introduction

In Indian manufacturing environments, according to the IFR World Robotics Report 2024, India is one of the countries where growth in robotics has been quite remarkable. In 2023, 8,510 industrial robots were installed in India, which saw a 59% increase from the previous year and made India the 7th largest market for annual robot installations in the world. The industry that led the charge was the automobile, with a jump of 139% to 3,551 units, which is shown in Figure 1, and this accounted for 42% of the market. In today’s integration of cobots in India, most of them are fixed to one position and dedicated to one task in the production line. Due to the lack of mobility and flexibility, there are several challenges.^1,2 One of them is that due to a lack of mutual dependency, production time increases. If any malfunctions occur, there are a lot of potential lost hours, and even large-scale production at particular workstations may stop.³ These cobots are unable to dynamically reallocate themselves; the inefficiency impacts overall productivity and increases the fragility of production workflows in times of disruptions. Therefore, an enhanced system for collaborative robots is needed to better optimize production workflows and decrease downtime.^4–6

The primary objective of this paper is to highlight the potential of the AGV-based cobot in manufacturing environments through simulation studies. The current work focuses on demonstrating navigation, docking, and task execution within a controlled virtual setup. While aspects such as dynamic human interactions, environmental changes, and unexpected disruptions are important, they are beyond the scope of this simulation-based study. These factors will be carefully explored in future real-world implementations. The AGV-based cobot system demonstrates enhanced productivity and flexibility compared to conventional fixed cobots in modern manufacturing environments. Unlike their static counterparts, AGV-based cobots exhibit dynamic mobility, allowing them to traverse between multiple workstations and adapt to layout changes without the need for reinstallation. This mobility significantly reduces cycle time per task, achieving completion rates of 12–15 seconds (including transit time), whereas fixed cobots require 18–22 seconds, resulting in a 20–30% faster task completion rate. Moreover, AGV-cobots minimize idle time by autonomously moving to active stations, unlike fixed cobots, which suffer from higher idle periods due to their limited reach. This dynamic movement results in 25–40% lower idle time, enhancing overall operational efficiency.^7,8

The multitasking capabilities of AGV-based cobots further boost productivity by enabling them to execute tasks at multiple stations, whereas fixed cobots are restricted to a specific location.⁹ This flexibility leads to a 25–35% increase in production throughput. Additionally, AGV-based cobots optimize resource utilization by dynamically reallocating tasks, making them more efficient in comparison to the static task allocation of fixed cobots. The system also demonstrates superior space optimization, requiring up to 30% less dedicated workspace by eliminating the need for fixed infrastructure. During layout reconfigurations, AGV-based cobots exhibit minimal downtime, whereas fixed cobots experience substantial delays due to reinstallation requirements. The scalability of AGV-based cobots is another advantage, as they can be easily expanded by integrating additional units, while fixed cobots offer limited scalability due to their static setup. Overall, AGV-based cobots significantly enhance productivity, resource utilization, and operational flexibility, making them a more efficient solution for dynamic and adaptive manufacturing systems.

This work provides an alternative solution to overcome the drawbacks. In this study, the cobot is designed with flexibility as it has a changeable end effector with AGV, thus it can perform different processes at separate workstations. It moves from one workstation to another, mounted on an Autonomous Guided Vehicle (AGV).^10,11 Moreover, the Dynamic Integration of Collaborative Robots and Automated Guided Vehicles in Manufacturing Systems provides active and passive modes of operation for easy human and cobot collaboration.¹² A simulation of the Dynamic Integration of Collaborative Robots and Automated Guided Vehicles in Manufacturing Systems demonstrates how this can be effectively operational across four different workstations within a diesel engine manufacturing facility. By considering 4 workstations, where each workstation was connected to a network of AGV-based cobots. These cobots can interchange throughout the workstations and can change the end effector.

The workstations have fixed environments that are dedicated to a particular assigned task. A network of AGV-based cobots can do the work at the assigned workstation. The cobots that are fully functional and idle can assist the other cobots as well as humans in other workstations. The cobots can exchange the work so that the overload of work on a single cobot can be reduced, which increases Overall equipment effectiveness (OEE). Implementing AI can increase the quality of the process and produce a seamless operation. These cobots are connected through a centralized database where humans can access the cobots’ network through the Human Machine Interface (HMI).^13,14

Methodology

The objective of this idea is to make an AGV-based cobot that can do work at any time, anywhere, and in any process. By developing it in multiple workstations of a diesel engine manufacturing unit. Figure 2. Explains the detailed process flow of the idea. The simulations for this work were performed using CoppeliaSim v4.1.0, a simulation platform for robotics and automation. It is known for its real-time simulation capabilities. In this study, the software was used to design and validate the control algorithms for multiple AGV-based cobots. Its ability to integrate with external controllers and simulate sensor dynamics allowed for precise testing of navigation and obstacle avoidance strategies. The choice of CoppeliaSim was based on its adaptability and realistic modeling, which ensured reliable results that complied with real-world conditions.

The AGV-based cobot system in CoppeliaSim employs a line-following algorithm for navigation, ensuring the AGV accurately tracks predefined paths using visual cues. The robot’s motion planning relies on Forward Kinematics (FK) and Inverse Kinematics (IK) using a traditional analytical approach. Forward kinematics (FK) computes the end-effector’s pose from given joint angles, while inverse kinematics (IK) determines the joint angles required to reach a desired pose. Standard solutions consider the robot’s degrees of freedom and workspace constraints.¹⁵ To improve trajectory accuracy, the system dynamically fine-tunes the joint angles using real-time feedback, ensuring smooth motion and effective collision avoidance. Since the simulation assumes a fixed workspace with no dynamic obstacles, the path remains stable and predictable. The AGV operates at a maximum speed of 0.2 ms–1, with controlled acceleration and deceleration to prevent overshooting its payload capacity of up to 10 kg. The simulation parameters, including joint limit constraints, workspace boundaries, and trajectory smoothing techniques, ensure realistic and reliable performance, making the system well-suited for dynamic manufacturing environments.

The CoppeliaSim simulation platform was chosen for this study due to its powerful physics engine, which accurately models the dynamic and kinematic behavior of AGV-based cobots. It offers integrated APIs for Python, Lua, and C++, enabling the implementation of custom control algorithms and flexible interaction with the simulation environment. CoppeliaSim also supports real-time control, collision detection, and multi-threaded simulation, making it suitable for testing cobot operations in complex manufacturing scenarios. Compared to other simulators, Webots offers high-fidelity rendering and realistic physics for robot perception and environment visualization, but it has fewer built-in control features.^16,17 On the other hand, Gazebo is widely used for its integration with ROS (Robot Operating System), making it ideal for real-world deployment testing, although its steeper learning curve and higher computational requirements can hinder rapid prototyping. Despite its strengths, CoppeliaSim has limitations in real-world applicability, as it does not fully capture practical issues such as sensor noise, wheel slippage, or unpredictable human interactions. Therefore, while it is effective for design validation and algorithm testing, real-world testing is necessary to verify the system’s performance under practical operating conditions.

The Environment of each workstation is assigned to a specified work process, such as workstation 1 is for drilling and deburring, workstation 2 is for inspection, workstation 3 is for assembly and fastening, and workstation 4 is for coating and painting. Cobot is further supported by its components like teaching pendants, PLC, controller, battery, sensors, and vision systems.18 These parts ensure that cobots can adapt to several tasks with high precision. The cobot has interchangeability of end effectors and features grinding, drilling, fastening, pick-and-place, and inspection tools. This gives the ability for the cobot to easily switch between tasks between workstations. The navigation and stability are enhanced in the AGV by its guided lines and advanced sensors like torque, ultrasonic, proximity, GPS, and encoders, to achieve precise positioning and vibration-free processes.¹⁹

The manufacturing process flow incorporates rework or re-pair stations and a cobot maintenance station for any functional issues that could arise, thus reducing downtime and ensuring that the production process remains intact. The human-machine interface improves usability by allowing operators to interact with the cobot in terms of programming, monitoring, and troubleshooting. Such dynamic system design ensures that the cobot optimizes production processes, reduces costs on inventory, and maximizes flexibility and efficiency within the manufacturing unit. In CoppeliaSim, AGV communication uses TCP/IP-based remote API calls, ensuring real-time data exchange for navigation and task execution. This enables low-latency control and efficient coordination in the simulation environment. However, in real-world applications, protocols such as MQTT or ROS over TCP/IP can be commonly used, offering reliable and scalable communication. While CoppeliaSim ensures seamless data transfer, real-world networks face latency and congestion, impacting system efficiency. Therefore, the protocol choice plays a vital role in ensuring robust and responsive AGV coordination.

Collaborative Robot Design

Integration of Automated Guided Vehicle: The AGVs are integrated into the collaborative robot to allow mobility, so the cobot can work at multiple workstations, decrease the production time, and increase flexibility. The navigation is ensured through guided lines and sensors in the AGVs for accurate alignment at the desired workstation. The cobot is mounted on a stable AGV platform to minimize vibrations during operations, ensuring accuracy and reliability. Guided pathways or pre-mapped routes help the AGV travel efficiently across the manufacturing floor.²⁰ Connectivity Between All the AGVs: With connectivity among many AGVs, it is able to coordinate them to find their target workstation, which then moves accordingly. The AGVs communicate among themselves with the use of a centralized control system, and the machine learning algorithms will ensure smooth workflow in every station. In this approach, Radio frequency identification (RFID) tags at strategic locations on the workstation help the navigation of AGVs based on location information.^21,22 It is also applied to predict paths, avoid collisions, and allocate AGVs to various tasks efficiently to enhance the overall system.

Collaborative Robot Manipulator Design: The collaborative robot manipulator employs a 6-degree-of-freedom structure integrated with the AGV system, enabling precise positioning and flexible task execution across multiple workstations, which is a referred model of KUKA LBR iiwa 14 R820 cobot.²³ Machine learning techniques are used to enhance motion stability, optimize trajectory planning, and support interchangeable end effectors, thereby improving adaptability and operational efficiency under varying workspace constraints.

Table 1 presents a comparative summary between the fixed-base cobot and the AGV-based cobot in terms of workspace, cycle time, performance, and energy efficiency, highlighting the estimated improvements achieved through mobility integration. The estimated improvement demonstrates that the mobile cobot provides substantial gains over the fixed base system. Specifically, the mobile cobot expands the effective workspace by approximately 200–300%, enabling it to serve multiple stations without manual part transport, whereas the fixed cobot is limited to a single workstation. In terms of productivity, mobile operation allows manipulation tasks to be performed while moving, significantly reducing waiting times. This results in up to 48% reduction in task completion time compared to the fixed-base cobot. Moreover, the combination of mobility and manipulation precision leads to overall system throughput improvements of 25–35%, as relocation eliminates delays associated with human or conveyor transport. Energy-wise, while the mobile base consumes additional power, the reduction in manual transport or conveyor usage provides a net energy/time saving of 10–20% in typical tasks. These results quantitatively substantiate the claimed performance gains of 20–40% for mobile cobots over fixed-base systems, highlighting improvements in workspace coverage, cycle time, throughput, and energy efficiency, and confirming the added operational flexibility provided by mobile manipulators.

Table 1: Comparison table between fixed-based cobot and AGV-based cobot.
Attribute	Fixed Base Cobot	AGV-Based Cobot	Estimated Improvement
Workspace	Works only at one fixed spot; parts must be brought to it.	Can move around, reach multiple spots, and cover a wider area.	~200–300% more coverage area.
Cycle Time	Can work fast, but delays happen while waiting for parts, people, or transport.	Can keep moving while working; one study showed up to 48% faster task completion.	~48% faster in that case.
Performance	Measured only for the robot arm; well-established benchmarks.	Measured on time, distance, completion, and effect of moving base; keeps good precision, slightly less repeatable, but overall throughput improves.	~25–35% better throughput.
Energy Efficiency	Uses energy only for the arm (no movement).	Uses energy for both arm and movement; may save energy/time by reducing human or conveyor transport.	~10–20% savings in some cases.

End Effector Locking Mechanisms

To switch end effectors quickly and safely, various locking mechanisms are discussed, including pneumatic, hydraulic, magnetic, and mechanical systems. Facilitate rapid and safe end-effector changeovers, the magnetic locking mechanism has been chosen as the basic system for its simplicity, reliability, and efficiency. The magnetic system is appropriate for attaching and detaching end effectors securely with little effort and minimal downtime. The clamping force provided by this system is sui for many manufacturing processes, with sufficient ease of use and operating speed. Standardization of magnetic locking mechanisms throughout the system leads to smooth, uniform, and consistent. Changes in the end effector result in increased flexibility and efficiency in application.^24–26

The magnetic locking mechanism has been validated for payloads up to 10 kg in line with ISO/TS 15066 safety standards. Redundant current feedback sensors and emergency-stop protocols ensure compliance with human-robot collaboration safety requirements. The mechanism, integrated into the AGV-based cobot, employs high-strength electromagnets with sufficient holding force to ensure secure and reliable docking of the end effector, preventing unintentional detachment even during high-speed movements or minor collisions. As illustrated in Figure 3 and Figure 4, the magnetic locking of the end effector enhances safety by reducing the risk of payload drops or misalignments. Furthermore, the system incorporates power-fail retention to maintain the locked state during brief power losses and redundant power monitoring to prevent accidental disengagement, thereby improving operational reliability in dynamic manufacturing environments. Additionally, current feedback monitoring detects anomalies, ensuring consistent locking performance and enhancing overall safety.²⁵

Changeable End Effectors for Operations

Drilling: At workstation 1, the base of the diesel engine body must be drilled as per the required sections. The drilling end effector is designed to handle all kinds of materials and ensure accuracy in drilling operations. Various drilling end effectors, for example, adjustable speed and specific drill bit geometries, are analyzed to ensure the requirements of diesel engine manufacturing. The selection process will involve material compatibility, accuracy, and maintenance requirements.

Deburring: Deburring tools are designed to efficiently remove sharp edges and burs left behind from machining processes. The options include rotary deburring tools, abrasive pads, and brush systems that are analyzed based on their capability to deal with different surface profiles and material types. The chosen end effector ensures a smooth finish while minimizing cycle time. This end effector is used at workstation 1 to remove the sharp edges and burs left.

Inspection: In workstation 2, this end effector is used for the inspection of diesel engine parts. Ensuring quality control in the manufacturing process. Techniques involving AI and ML have been included to enhance the detection of defects and aid in decision-making while inspection is carried out. This end effector can take high-resolution images, perform dimensional inspections, and find surface defects with great precision.²⁷

Pick and Place: Grippers are the end effectors that are used in an assembly section to assemble the diesel engine. The pick-and-place end effectors are material handling components. These are equipped with mechanisms such as vacuum suction, grippers, or magnetic claws. Such systems are examined to check their capability for processing different components of all shapes, sizes, and weights at the same level of speed and accuracy.

Painting: The painting end effector is utilized in a separate workstation so that the changing of the manipulators only belongs to the paint station. Paint end effectors are designed for uniform coating and precise application of paint on components. Options such as spray nozzles, roller systems, or electrostatic applicators are evaluated based on paint type, surface area, and desired finish quality. These end effectors ensure efficiency in coating and painting processes while minimizing wastage.²⁸

Control Methods

Active mode: In active mode, the cobot is controlled using a teach pendant and human-assisted positioning systems.29,30 The active mode allows for direct control of the movements of the cobot, so in case of complex or accuracy-sensitive tasks, the operators can guide the cobot manually. The teaching pendant facilitates user-friendly programming and monitoring. Human-assisted positioning systems enhance adaptability by allowing operators to intervene in real-time and adjust the cobot’s position, especially in tasks needing fine adjustments or interaction with delicate components. This is highly effective in dynamic manufacturing environments, where one needs precision and responsiveness to operations.

Passive mode: In passive mode, the cobot can be guided and taught by manual movement of the operator to various desired positions. Once a teaching cycle is done, the cobot is competent enough to continue repeating the operation without additional intervention until the operator decides that the process or task must be changed. This contact-oriented teaching makes operations relatively simpler and, therefore, allows even minimal technical know-how operators to get started with work. The passive mode allows for easy human-cobot collaboration. It allows operators to intuitively train the cobot for new tasks without the use of advanced programming tools. The mode increases flexibility in operation since the cobot can quickly adapt to different tasks or workstations, making it particularly ideal for environments where process changes are frequent or customized production requirements are applied.

Path Planning Algorithm of an Automated Guided Vehicle

Navigation of the AGV-based cobot in CoppeliaSim was implemented using a reactive line-following algorithm. Red lines on the virtual factory floor served as guides, and the AGV tracked them in real time using simulated vision sensors, continuously adjusting its heading without precomputing the full path as illustrated in Figure 5 and Figure 6. The AGV was modeled as a two-wheeled differential drive vehicle with non-holonomic constraints, assuming pure rolling without lateral slip. Upon reaching a destination, the cobot executed pre-programmed tasks, such as pick-and-place operations. End-effector motion was governed by forward and inverse kinematics, with forward kinematics computing the pose from joint angles and inverse kinematics determining the joint positions required to reach target positions. Real-time feedback control adjusted wheel speeds and joint angles to correct minor deviations during motion, while virtual proximity sensors enabled collision avoidance by triggering corrective maneuvers or controlled stops. The simulation employed a 50 ms time-step, a wheel–floor friction coefficient of 0.7, and Gaussian sensor noise (σ = 0.02 m) to approximate realistic navigation uncertainties.

Simulation

The simulation of combining collaborative robots with automated guided vehicles in a manufacturing system shows how AGVs can work with cobots across different workstations in a diesel engine factory. Each workstation has specific tasks for the cobot, and the system uses advanced control and changeable end effectors to get better productivity and accuracy. The full factory arrangement is shown in the layout as illustrated in Figure 7, and the KUKA LBR iiwa 14 R820 cobot used in CoppeliaSim V 4.1.0 is shown in Figure 8. The simulation was run using the technical details of the KUKA cobot, such as 14 kg payload, 820 mm reach, and its joint speed and torque limits.²³ Incorporating these properties into the model ensures that the simulation closely reflects the expected performance of a real system. Although absolute cycle times, energy consumption, and utilization may differ slightly due to factors such as sensor noise, wheel slippage, and mechanical wear, the observed trends improved throughput and reduced idle time compared to fixed cobots or manual operation remain consistent and reliable.

Workstation 1 Drilling and Deburring

Workstation 1 is specifically assigned to deburring and drilling, as shown in Figure 9. In this, one cobot is mounted on an AGV and operated by an operator through a teach pendant. This allows the cobot to be positioned precisely and the task to be executed accurately. Meanwhile, another cobot arrives on another AGV to perform parallel tasks, thereby increasing productivity. The deburring cobot smoothes the rough edges of the mold and components by using a rotary deburring tool, whereas the drilling cobot operates with adjustable speed and specialized drill bits for creating holes in different materials. This collaborative and parallel operation reduces cycle time, increases flexibility, and ensures high-quality outcomes for the diesel engine components.

Workstation 2 Inspection

In Workstation 2, quality control through advanced inspection is the focus shown in Figure 10. Two AGVs, each mounted with a cobot, reach the station. These cobots are mounted with vision sensors and AI-enabled inspection end effectors. They detect surface defects and check if the components meet the standards of quality through high-resolution imaging and dimensional checks. Only the External part of the body is inspected by the cobots, and the internal inspection is done by different methods by humans. Advanced image processing techniques based on AI further enhance the ability of the system to find defects. The cobots work in tandem to inspect various aspects of the parts, using real-time data to make accurate determinations. The addition of machine learning algorithms helps predict and analyze defects, which further enhances the reliability and consistency of the inspection process. This configuration not only improves the quality control stage but also reduces potential rework downstream.

Workstation 3 Assembly

At Workstation 3, the cobots assemble the diesel engine shown in Figure 11. In this workstation, pistons, crankshafts, and blocks of the engine are assembled with the help of interchangeable end effectors that allow for fastening and mechanical alignment. The AGV-mounted cobots ensure that transitions between subassemblies occur without any interruptions in the flow, while still maintaining precision torque application and alignment. The diesel engines assembled here are then transferred to AGVs and moved on to the next stage. The system has reduced interference from human beings since it automates the assembly process, saves time on assembly, and ensures consistency in quality on each unit. This ability of cobots in the workstation demonstrates the effectiveness of the 3A system in complicated manufacturing operations.

Workstation 4 Painting

The final production operation occurs at Workstation 4, shown in Figure 12, where assembled diesel engines are painted and treated before proceeding to subsequent processes. The AGV-based cobot at this station operates independently of other workstations. Its painting end effector employs a spray nozzle system for uniform coating and precise application, with parameters such as paint thickness and surface finish monitored in real time to minimize waste and ensure quality. AI-driven quality checks further ensure that the process meets stringent standards. After painting, engines undergo a curing stage before being shipped or stored. Integrating cobots at this station streamlines finishing operations, enhances consistency, and reduces processing time by eliminating manual adjustments. The simulation demonstrates how the collaborative robot can traverse multiple workstations, highlighting its potential to reduce production times, minimize downtime, and lower manufacturing costs in diesel engine assembly.

Results and Discussion

The integration of collaborative robots (cobots) and automated guided vehicles (AGVs) presents a promising approach to enhancing manufacturing operations, particularly in diesel engine assembly. Dynamic reallocation of cobots significantly reduces idle time and maximizes resource utilization, unlike stationary systems, where inefficiencies are inevitable. The current study presents preliminary simulation results for the conceptual design of an AGV cobot system developed in CoppeliaSim. Quantitative data were collected over 14 independent simulation runs, capturing cycle time and production rate for each workstation. Across the simulations, the mean cycle times were 73.29 ± 1.86 s for Workstation 1, 86.07 ± 1.49 s for Workstation 2, 118.07 ± 1.73 s for Workstation 3, and 63.50 ± 1.09 s for Workstation 4. Correspondingly, the mean production rates achieved were 48.79 ± 1.31, 41.29 ± 0.83, 30.07 ± 0.62, and 56.50 ± 1.09 units per hour, respectively. While precise real-world validation against manual operation or fixed-cobot cells is not feasible at this conceptual stage, these simulation outcomes provide indicative benchmarks for throughput and workstation utilisation under idealised conditions. The results suggest the potential for performance improvements relative to conventional setups, particularly in terms of reduced cycle times, workspace, and enhanced production consistency, which will be quantified against manual and fixed-cobot baselines in future stages of the work.

Simulation results further show that integrating cobots with AGVs in diesel engine production can lead to smoother and more efficient operations. Dynamic reallocation of cobots reduced idle time, optimized resource utilization, lowered costs, and increased throughput. Parallel operations at Workstation 1, as shown in Figure 5, shortened cycle times, while AI-aided inspections at Workstation 2, depicted in Figure 6, enabled accurate defect detection. Automated assemblies at Workstation 3, illustrated in Figure 7, improved production rates, and painting operations at Workstation 4, presented in Figure 8, minimized material waste and ensured sustainable processing. These results highlight the potential of flexible cobot deployment to reduce production times, minimize downtime, and optimize costs.^31,32

The performance assessment in this study is primarily qualitative, based on demonstrative simulations conducted in CoppeliaSim. Improvements in efficiency compared to traditional fixed systems are evident through visualized task execution and workflow demonstrations, showing faster task completion, reduced idle time, and enhanced operational flexibility. As the study does not include detailed quantitative metrics or real-world trials, these results serve as a proof of concept to illustrate the system’s potential in dynamic manufacturing environments. Real-world validation is necessary to quantify these benefits, taking into account factors such as sensor inaccuracies, dynamic obstacles, and environmental uncertainties, which are not fully captured in simulation. The current setup does not explicitly model equipment failures or human interactions, although the AGV-based cobot is capable of handling unexpected obstacles through real-time path adjustments using its trajectory tracking algorithm. For practical deployment, integrating fault detection protocols and human safety mechanisms, including emergency stops and collision avoidance, will be essential to ensure safe and reliable operation, as given in Table 2.^33,34

Table 2: Limitations and impact on AGV-based cobot system.
Limitation	Impact on AGV-Cobot System	Mitigation/Proposed Solution	Supporting Studies
Sensor drift/localization errors	Cumulative positioning errors reduce navigation accuracy	Sensor fusion (odometry + IMU + laser scans), extended Kalman filter	Mobile manipulator localization studies36
Wheel slippage	Path tracking deviations, reduced mobility performance	Adaptive and robust control algorithms, slippage compensation	Mecanum-wheeled robot control37
Human unpredictability	Variability in collaborative tasks, potential safety hazards	Intuitive interfaces, dynamic human-robot collaboration protocols, predictive motion planning	Human-scale mobile manipulators38
Communication latency / wireless delays	Delayed control feedback affecting task coordination	Optimized wireless protocols, edge-computing integration	Industrial mobile robot networks39
Power failures	Sudden stops or system downtime	Uninterruptible power supply, fail-safe braking	Standard AGV safety practices40
Maintenance challenges	Reduced operational uptime	AI-driven predictive maintenance, scheduled preventive maintenance	Industrial robotics maintenance frameworks41

Scalability of the proposed system to larger and more complex manufacturing environments is feasible, as simulations indicate that multiple AGV-cobot units can operate concurrently and coordinate via centralized communication protocols.35 The modular design, including interchangeable end effectors for tasks such as drilling, inspection, and painting, allows rapid adaptation to varying manufacturing requirements without extensive reconfiguration. Machine learning algorithms facilitate real-time communication, navigation, collision avoidance, and task allocation among AGVs. Simulations show that the system can handle dynamic layouts, multi-tasking, and adaptive path planning, suggesting suitability for larger-scale deployments. However, as these findings are based solely on simulation, real-world validation is necessary to confirm practical scalability, robustness, and operational efficiency.

The modular design of the system, combined with a centralized control network, enhances both scalability and adaptability. Interchangeable end effectors for tasks such as drilling, inspection, and painting allow cobots to adapt quickly to varying manufacturing requirements without extensive reconfiguration. AI and machine learning algorithms support real-time communication and coordination among AGVs, enabling efficient navigation, collision avoidance, and dynamic task allocation. While these capabilities are promising, the results are based solely on simulation and do not fully capture challenges present in unstructured production floors, unexpected equipment failures, or human interactions. Specific limitations such as sensor drift, wheel slippage, communication latency, power failures, and maintenance challenges are summarized in Table 2, along with their impacts on the AGV-cobot system and potential mitigation strategies supported by prior studies. Future work will involve testing under variable physical conditions to evaluate robustness, reliability, and large-scale industrial potential, including real-world trials in collaboration with industry partners.

The proposed AGV-based cobot system introduces an innovative approach to flexible and adaptive manufacturing, demonstrating promising results in simulation environments. However, we acknowledge that detailed experimental validation and real-world testing are necessary to confirm its practical feasibility and effectiveness. Future work will focus on collaborations with industry partners to conduct real-world trials and assess the system’s performance under actual manufacturing conditions. Additionally, more comprehensive comparisons with existing technologies, particularly in terms of scalability, efficiency, and reliability in industrial applications, will be conducted to further substantiate the system’s advantages and applicability.

While the simulation study demonstrates the feasibility of the AGV-based cobot system, certain limitations must be acknowledged when considering real-world deployment. In practical manufacturing environments, factors such as wheel slippage can reduce navigation accuracy, particularly on uneven or high-friction surfaces. Similarly, localisation drift may occur over prolonged operations due to cumulative sensor errors, requiring advanced correction strategies such as sensor fusion or external calibration. Safety and compliance considerations are central to real-world deployment. The system is designed according to ISO 10218 for industrial robots and ISO 3691–4 for automated guided vehicles. Wireless communication latency is optimized to remain below 50 ms for real-time task coordination, while power-fail contingencies are addressed using uninterruptible power supplies and fail-safe braking mechanisms. Furthermore, payload constraints may limit the versatility of interchangeable end-effectors, as each attachment must remain within defined load and stability margins to ensure safe operation. To bridge the gap between simulation and real-world applicability, a future hardware validation plan is proposed, involving physical prototyping, safety testing under ISO guidelines, and iterative evaluation of navigation, load-handling, and human–robot interaction in dynamic factory environments.

Conclusion

The Dynamic Integration of Collaborative Robots and Automated Guided Vehicles in Manufacturing Systems presents a paradigm in manufacturing automation through mobility, versatility, and intelligence.⁴² Simulation studies in a diesel engine manufacturing facility demonstrate that AI- and ML-driven AGV-based cobot systems can significantly improve productivity, product quality, and cost efficiency. This can automate the entire production process and reduce the requirement for skilled employees. Such systems have the potential to automate complex production processes while reducing reliance on highly skilled labor, creating opportunities for unskilled and semi-skilled workers to participate effectively through intuitive interfaces and user-friendly programming. By implementing these systems in India, manufacturing operations can achieve higher throughput while broadening workforce inclusion.

Beyond technical benefits, this integration enhances human-cobot collaboration, allowing operators with minimal training to engage effectively in production roles. Automation of intricate processes, combined with seamless human-cobot interaction, positions the system as a critical enabler for modernizing manufacturing facilities, particularly in regions where robotics adoption is still expanding. Simulation results show improvements in mobility, reduced idle time, and optimized resource utilization compared to traditional fixed cobots. Despite these promising outcomes, real-world deployment may present challenges, including equipment compatibility, safety compliance, communication reliability, path deviations, unexpected human interactions, and maintenance requirements. Future research should focus on testing realistic manufacturing environments to study human intervention, environmental variability, and operational disruptions. High-error handling combined with AI-driven decision optimization can further improve system reliability, energy efficiency, and sustainability, supporting Industry 4.0-compatible smart production practices.

The results presented here are based on simulation studies conducted in CoppeliaSim, providing proof of concept for navigation accuracy, docking efficiency, and task execution. While real-world validation has not yet been conducted, planned collaborations with industry partners will allow experimental testing to confirm scalability, robustness, and practical effectiveness. Overall, the proposed AGV-based cobot system offers a promising framework for flexible, efficient, and inclusive manufacturing, bridging automation with workforce participation in a modern industrial context.

References

1. Castillo JF, Hamilton Ortiz J, Díaz Velásquez MF, Saavedra DF. COBOTS in Industry 4.0: Safe and Efficient Interaction. In: Collaborative and Humanoid Robots. London: IntechOpen; 2021. https://doi.org/10.5772/intechopen.101508
2. Keshvarparast A, Battini D, Battaia O, Pirayesh A. Collaborative robots in manufacturing and assembly systems: literature review and future research agenda. J Intell Manuf. 2024;35(5):2065–118. https://doi.org/10.1007/s10845-023-02026-3
3. Ojstersek R, Buchmeister B, Javernik A. The importance of cobot speed and acceleration on the manufacturing system efficiency. Procedia Comput Sci. 2023;217:147–54. https://doi.org/10.1016/j.procs.2023.02.090
4. Correia Simões A, Lucas Soares A, Barros AC. Factors influencing the intention of managers to adopt collaborative robots (cobots) in manufacturing organizations. J Eng Technol Manag. 2020;57:101574. https://doi.org/10.1016/j.jengtecman.2020.101574
5. Taesi C, Aggogeri F, Pellegrini N. COBOT applications—recent advances and challenges. Robotics. 2023;12(3):79. https://doi.org/10.3390/robotics12030079
6. Aaltonen I, Salmi T, Marstio I. Refining levels of collaboration to support the design and evaluation of human–robot interaction in the manufacturing industry. Procedia CIRP. 2018;72:93–8. https://doi.org/10.1016/j.procir.2018.03.162
7. Amiri P, Müller M, Southgate M, Theodoridis T, Wei G, Richards-Brown M, et al. A statistical analysis of commercial articulated industrial robots and cobots. J Manuf Mater Process. 2024;8(5):216. https://doi.org/10.3390/jmmp8050216
8. International Federation of Robotics. Record of 4 million robots in factories worldwide. Frankfurt: IFR; 2025 [cited 2025 Aug 26]. Available from: https://ifr.org/ifr-press-releases/news/record-of-4-million-robots-working-in-factories-worldwide
9. Javaid M, Haleem A, Singh RP, Rab S, Suman R. Significant applications of cobots in the field of manufacturing. Cogn Robot. 2022;2:222–33. https://doi.org/10.1016/j.cogrob.2022.09.004
10. De Ryck M, Versteyhe M, Debrouwere F. Automated guided vehicle systems, state-of-the-art control algorithms and techniques. J Manuf Syst. 2020;54:152–73. https://doi.org/10.1016/j.jmsy.2019.12.002
11. D’Souza F, Costa J, Pires JN. Development of a solution for adding a collaborative robot to an industrial AGV. Ind Robot. The international journal of robotics research and application. 2020;47(5):723–35. https://doi.org/10.1108/IR-03-2020-0047
12. Matheson E, Minto R, Zampieri EGG, Faccio M, Rosati G. Human–robot collaboration in manufacturing applications: a review. Robotics. 2019;8(4):100. https://doi.org/10.3390/robotics8040100
13. Moor M, Sarkans M, Kangru T, Otto T, Riives J. AI functionalities in cobot-based manufacturing for performance improvement in quality control application. J Mach Eng. 2024;24(2):44–55. https://doi.org/10.36897/jme/181078
14. Manju Bargavi SK, Tejashvar KR, Tejashwini KR. Artificial intelligence–driven cobots for innovative Industry 5.0 workforce. In: Intelligent Robots and Cobots. Hoboken: Wiley; 2025. p.159–75. https://doi.org/10.1002/9781119912871.ch8
15. Oswal S, Saravanakumar D. Line following robots on factory floors: significance and simulation study using CoppeliaSim. IOP Conf Ser Mater Sci Eng. 2021;1012(1):012008. https://doi.org/10.1088/1757-899X/1012/1/012008
16. Michel O. Webots: professional mobile robot simulation. J Adv Robot Syst [Internet]. 2004;1(1):39–42. Available from: http://www.ars-journal.com/International-Journal-of-
17. Koenig N, Howard A. Design and use paradigms for Gazebo, an open-source multi-robot simulator. In: IEEE/RSJ Int Conf Intell Robots Syst. Sendai, Japan: IEEE; 2004. p.2149–54. https://doi.org/10.1109/IROS.2004.1389727
18. Hameed A, Ordys A, Możaryn J, Sibilska-Mroziewicz A. Control system design and methods for collaborative robots: review. Appl Sci. 2023;13(1):675. https://doi.org/10.3390/app13010675
19. dos Reis WPN, Morandin Junior O. Sensors applied to automated guided vehicle position control: a systematic literature review. Int J Adv Manuf Technol. 2021;113(1–2):21–34. https://doi.org/10.1007/s00170-021-06598-8
20. Bhargava A, Suhaib M, Singholi AS. A review of recent advances, techniques, and control algorithms for automated guided vehicle systems. J Braz Soc Mech Sci Eng. 2024;46(7):419. https://doi.org/10.1007/s40430-024-04224-0
21. Isrofi A, Setiawan B, Wibowo S. Automated guided vehicle navigation control using matrix method applying radio frequency identification point address. In: 2021  International Conference on Electrical and Information Technology (IEIT). IEEE; 2021. p.139–46. https://doi.org/10.1109/IEIT53149.2021.9584241
22. Lu S, Xu C, Zhong RY. An active RFID tag-enabled locating approach with multipath effect elimination in AGV. IEEE Trans Autom Sci Eng. 2016;13(3):1333–42. https://doi.org/10.1109/TASE.2016.2549039
23. KUKA LBR iiwa 14 R820 [Internet]. [cited 2025 Aug 26]. Available from: https://my.kuka.com/s/product/lbr-iiwa-14-r820/01t58000002hnktAAA
24. Dai Y, Xiang C, Qu W, Zhang Q. A review of end-effector research based on compliance control. Machines. 2022;10(2):100. https://doi.org/10.3390/machines10020100
25. Iqbal Z, Pozzi M, Prattichizzo D, Salvietti G. Detachable robotic grippers for human–robot collaboration. Front Robot AI. 2021;8:708396. https://doi.org/10.3389/frobt.2021.708396
26. Feng F, Tang L, Xu J, Liu H, Liu Y. A review of the end-effector of large space manipulator with capabilities of misalignment tolerance and soft capture. Sci China Technol Sci. 2016;59(11):1621–38. https://doi.org/10.1007/s11431-016-6019-9
27. Somwang P, Muangklang E. Image processing for quality control in manufacturing process. In: 2019 16th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON). IEEE; 2019. p. 782–5. https://doi.org/10.1109/ECTI-CON47248.2019.8955259
28. Muzan IW, Faisal T, Al-Assadi HMAA, Iwan M. Implementation of industrial robot for painting applications. Procedia Eng. 2012;41:1329–35. https://doi.org/10.1016/j.proeng.2012.07.331
29. Dong J, Kwon W, Kang D, Nam SW. A study on the usability evaluation of teaching pendant for manipulator of collaborative robot. In: 2021. p.234–8. https://doi.org/10.1109/IIAI-AAI53464.2021.00042
30. Faccio M, Granata I, Menini A, Milanese M, Rossato C, Bottin M, et al. Human factors in cobot era: a review of modern production systems features. J Intell Manuf. 2023;34(1):85–106. https://doi.org/10.1007/s10845-021-01813-x
31. Weiss A, Wortmeier AK, Kubicek B. Cobots in Industry 4.0: a roadmap for future practice studies on human–robot collaboration. IEEE Trans Hum Mach Syst. 2021;51(4):335–45. https://doi.org/10.1109/THMS.2020.3045766
32. Vitolo F, Rega A, Di Marino C, Pasquariello A, Zanella A, Patalano S. Mobile robots and cobots integration: a preliminary design of a mechatronic interface by using MBSE approach. Appl Sci. 2022;12(1):419. https://doi.org/10.3390/app12010419
33. Soori M, Arezoo B, Dastres R. Artificial intelligence, machine learning and deep learning in advanced robotics: a review. Cogn Robot. 2023;3:54–70. https://doi.org/10.1016/j.cogrob.2023.02.002
34. El Zaatari S, Marei M, Li W, Usman Z. Cobot programming for collaborative industrial tasks: an overview. Rob Auton Syst. 2019;116:162–80. https://doi.org/10.1016/j.robot.2019.03.005
35. Rigó L, Fabianová J, Palinský J, Dočkalíková I. Simulation and optimization of an intelligent transport system based on freely moving automated guided vehicles. Appl Sci. 2024;14(17):7937. https://doi.org/10.3390/app14177937
36. Ghodsian N, Benfriha K, Olabi A, Gopinath V, Arnou A. Mobile manipulators in Industry 4.0: a review of developments for industrial applications. Sensors. 2023;23(19):8026. https://doi.org/10.3390/s23198026
37. Hendzel Z, Trojnacki M. Neural network control of a four-wheeled mobile robot subject to wheel slip. In: 2015. p.187–201. https://doi.org/10.1007/978-3-319-10380-8_15
38. Kessens CC, Kaplan M, Rocks T, Osteen PR, Rogers J, Stump E, et al. Human-scale mobile manipulation using RoMan. Field Robot. 2022;2:1232–62. https://doi.org/10.1016/j.cogrob.2022.09.004
39. Haxhibeqiri J, Jarchlo EA, Moerman I, Hoebeke J. Flexible Wi-Fi communication among mobile robots in indoor industrial environments. Mobil Inf Syst. 2018;2018:1–19. https://doi.org/10.1155/2018/3928703
40. Chen W, Liu G, Jiao D, Yang G, Wang R. A review of redundant power supply system based on autonomous vehicle. In: 2024 4th International Conference on Control Theory and Applications (ICoCTA). IEEE; 2024. p. 325–30. https://doi.org/10.1109/ICoCTA57932.2024.10181655
41. Pookkuttath S, Rajesh Elara M, Sivanantham V, Ramalingam B. AI-enabled predictive maintenance framework for autonomous mobile cleaning robots. Sensors. 2021;22(1):13. https://doi.org/10.3390/s22010013
42. Titu AM, Gusan V. The influence of collaborative robots on the quality, efficiency and effectiveness of automotive manufacturing flows. In: 2022. p.58–67. https://doi.org/10.1109/ICAMCM55688.2022.10080451.

Cite this article as:
Kuppan Chetty RM, Sailendra K, Pappula HK, Tharun T and Narayanan R. Dynamic Integration of Collaborative Robots and Automated Guided Vehicles In Manufacturing Systems – A Simulation Study. Premier Journal of Science 2025;15:100136

Export to EndNote Export to Mendeley