Maximilian Bega
Maximilian Bega has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Bernd Kuhlenkötter
Chair of Production Systems, Ruhr-University Bochum, Bochum, Germany
4 shared publications - Furkan Ercan
Ruhr-University Bochum,Chair of Production Systems,Bochum,Germany
3 shared publications - Christian Hopmann
Institute for Plastics Processing (IKV) in Industry and Craft, RWTH Aachen University, Seffenter Weg 201, 52074 Aachen, Germany
1 shared publication - Christopher Prinz
Faculty of Mechanical Engineering, Chair of Productionsystems, Ruhr-University Bochum, Bochum, Germany
1 shared publication - Enes Alp
Chair of Production Systems (LPS), Ruhr-Universität Bochum, Bochum, Germany
1 shared publication Hui Yuan
1 shared publication
Latest Publications
- Open Access
- 29 May 2026
- Published by Springer Nature in Production Engineering
- Vol. 20 (3)
Dynamic scheduling of human-robot collaboration with deep reinforcement learning
Dynamic scheduling in Human-Robot Collaboration (HRC) is crucial for realizing the human-centric manufacturing goals of Industry 5.0, yet it faces significant challenges from uncertainties, particularly those stemming from human factors, which impede real-time optimization. This paper proposes a novel dynamic scheduling model for HRC assembly environments, leveraging Deep Reinforcement Learning (DRL). The model utilizes a Graph Neural Network (GNN) to embed complex system states, including operator characteristics and task assignments (human-only, robot-only, and human-robot collaborative). A scheduling policy, trained using the Proximal Policy Optimization (PPO) algorithm, is developed to dispatch tasks dynamically with the objective of minimizing makespan while adapting to uncertain operational events. The efficacy of the proposed approach was evaluated through numerical experiments based on a realistic lithium battery pack assembly scenario, considering various HRC configurations and human performance variability modeled using skewed distributions. Comparative analysis against traditional metaheuristic algorithms (Genetic Algorithms with crisp, fuzzy, and neutrosophic number representations) and OR-Tools demonstrated the DRL model’s superior performance. It consistently achieved significantly lower makespan and improved scheduling stability (lower deviation in makespan), particularly in complex multi-human, multi-robot configurations and in managing process uncertainties. The findings highlight the potential of DRL to provide robust and adaptable solutions for dynamic HRC scheduling, thereby enhancing the efficiency and reliability of collaborative manufacturing systems.
- 29 May 2026
- Published by Springer Nature in Production Engineering
- Vol. 20 (3)
Open AccessDynamic scheduling of human-robot collaboration with deep reinforcement learning
Dynamic scheduling in Human-Robot Collaboration (HRC) is crucial for realizing the human-centric manufacturing goals of Industry 5.0, yet it faces significant challenges from uncertainties, particularly those stemming from human factors, which impede real-time optimization. This paper proposes a novel dynamic scheduling model for HRC assembly environments, leveraging Deep Reinforcement Learning (DRL). The model utilizes a Graph Neural Network (GNN) to embed complex system states, including operator characteristics and task assignments (human-only, robot-only, and human-robot collaborative). A scheduling policy, trained using the Proximal Policy Optimization (PPO) algorithm, is developed to dispatch tasks dynamically with the objective of minimizing makespan while adapting to uncertain operational events. The efficacy of the proposed approach was evaluated through numerical experiments based on a realistic lithium battery pack assembly scenario, considering various HRC configurations and human performance variability modeled using skewed distributions. Comparative analysis against traditional metaheuristic algorithms (Genetic Algorithms with crisp, fuzzy, and neutrosophic number representations) and OR-Tools demonstrated the DRL model’s superior performance. It consistently achieved significantly lower makespan and improved scheduling stability (lower deviation in makespan), particularly in complex multi-human, multi-robot configurations and in managing process uncertainties. The findings highlight the potential of DRL to provide robust and adaptable solutions for dynamic HRC scheduling, thereby enhancing the efficiency and reliability of collaborative manufacturing systems.
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Implementation of an OPC UA Information Model for KPI Calculations and Dashboard Reporting Based on ISO 22400
The advancement of industrial production benefits from the use of Key Performance Indicators (KPIs) for gaining a competitive advantage and guiding operational improvements. This contribution introduces an executable OPC UA Information Model (IM) in line with the ISO 22400-2 standard, which encompasses definitions for 38 KPIs and outlines their calculation from 50 data points. This new OPC UA IM's architecture enables the processing of KPI calculations using data directly obtained from machines or higher-level systems such as Enterprise Resource Planning (ERP) or Manufacturing Execution System (MES). Implemented in a industry-related lab environment, the IM demonstrates a solution for real-time KPI analysis and dashboard reporting. This paves the way for achieving advanced levels of operational visibility, in alignment with Industry 4.0 maturity milestones.
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Implementation of an OPC UA Information Model for KPI Calculations and Dashboard Reporting Based on ISO 22400
The advancement of industrial production benefits from the use of Key Performance Indicators (KPIs) for gaining a competitive advantage and guiding operational improvements. This contribution introduces an executable OPC UA Information Model (IM) in line with the ISO 22400-2 standard, which encompasses definitions for 38 KPIs and outlines their calculation from 50 data points. This new OPC UA IM's architecture enables the processing of KPI calculations using data directly obtained from machines or higher-level systems such as Enterprise Resource Planning (ERP) or Manufacturing Execution System (MES). Implemented in a industry-related lab environment, the IM demonstrates a solution for real-time KPI analysis and dashboard reporting. This paves the way for achieving advanced levels of operational visibility, in alignment with Industry 4.0 maturity milestones.
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Utilizing the Mixture-of-Agents Approach for Entity Resolution and Data Landscape Homogenization in Manufacturing Domains
The concept of Industry 4.0, despite its longstanding inception, continues to pose significant challenges for organizations striving for a uniform integration of technologies and processes within manufacturing environments. Many companies remain in the nascent stages of digital transformation, grappling with the standardization and harmonization of their digital infrastructure. This contribution explores the persistent issues related to the interoperability of assets and the integration of disparate digitization solutions that lack a cohesive framework. This contribution delves into the application a Mixture-of-Agents (MoA) approach to address Entity Resolution (ER) tasks in the manufacturing domain, aiming to align heterogeneous data landscapes with standardized models. However, integration of the MoA approach in an industry-oriented validation environment has revealed that it does not significantly enhance ER yet ($\Delta$ F1 $=6.67 \%$), although there is potential for future advancements.
- 12 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Utilizing the Mixture-of-Agents Approach for Entity Resolution and Data Landscape Homogenization in Manufacturing Domains
The concept of Industry 4.0, despite its longstanding inception, continues to pose significant challenges for organizations striving for a uniform integration of technologies and processes within manufacturing environments. Many companies remain in the nascent stages of digital transformation, grappling with the standardization and harmonization of their digital infrastructure. This contribution explores the persistent issues related to the interoperability of assets and the integration of disparate digitization solutions that lack a cohesive framework. This contribution delves into the application a Mixture-of-Agents (MoA) approach to address Entity Resolution (ER) tasks in the manufacturing domain, aiming to align heterogeneous data landscapes with standardized models. However, integration of the MoA approach in an industry-oriented validation environment has revealed that it does not significantly enhance ER yet ($\Delta$ F1 $=6.67 \%$), although there is potential for future advancements.
- Open Access
- 7 July 2023
- Published by Springer Nature in Production Engineering
- Vol. 17 (6) , 915-927
Extension of value stream mapping 4.0 for comprehensive identification of data and information flows within the manufacturing domain
The main driver of industry 4.0 is data and information for realizing resilient manufacturing systems through data-based decision-making. This requires the data flow from different machinery and its processing. Therefore, a methodical survey on the logical and technical requirements for enabling an efficient data flow can be beneficial since it fastens the process of establishing the data flow and gives production and software engineers a common understanding of the data’s meaning. Hence, data, related metadata like the unit or the frequency of recordings, and technical information, e.g., the machine’s interface or the communication protocol, are necessary. Based on the classical value stream mapping (VSM) as a lean management tool, the first approach for a methodical gathering of data, value stream mapping 4.0, was established in 2016. Since VSM 4.0 provides technical properties for specifying the data flow, some shortcomings were identified, particularly related to technical information. This contribution introduces the VSM 4.0 plus that solved the identified shortcomings by adding new properties. Compared to VSM 4.0, VSM 4.0 plus offers detailed specifications of data points, includes technical information, e.g., used communication protocols, adds information regarding data processing, and enhances the storage and usage dimension by providing more details. Through the added properties, VSM 4.0 plus allows domain and software engineers more efficient collaboration since all required information for establishing the data flow and its processing are united in one methodology. To verify the practicability, the VSM 4.0 plus was applied within several manufacturing-related use cases.
- 7 July 2023
- Published by Springer Nature in Production Engineering
- Vol. 17 (6) , 915-927
Open AccessExtension of value stream mapping 4.0 for comprehensive identification of data and information flows within the manufacturing domain
The main driver of industry 4.0 is data and information for realizing resilient manufacturing systems through data-based decision-making. This requires the data flow from different machinery and its processing. Therefore, a methodical survey on the logical and technical requirements for enabling an efficient data flow can be beneficial since it fastens the process of establishing the data flow and gives production and software engineers a common understanding of the data’s meaning. Hence, data, related metadata like the unit or the frequency of recordings, and technical information, e.g., the machine’s interface or the communication protocol, are necessary. Based on the classical value stream mapping (VSM) as a lean management tool, the first approach for a methodical gathering of data, value stream mapping 4.0, was established in 2016. Since VSM 4.0 provides technical properties for specifying the data flow, some shortcomings were identified, particularly related to technical information. This contribution introduces the VSM 4.0 plus that solved the identified shortcomings by adding new properties. Compared to VSM 4.0, VSM 4.0 plus offers detailed specifications of data points, includes technical information, e.g., used communication protocols, adds information regarding data processing, and enhances the storage and usage dimension by providing more details. Through the added properties, VSM 4.0 plus allows domain and software engineers more efficient collaboration since all required information for establishing the data flow and its processing are united in one methodology. To verify the practicability, the VSM 4.0 plus was applied within several manufacturing-related use cases.