Dean Brandner
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Dean Brandner has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Claudius Terkowsky
TU Dortmund University
3 shared publications - Konrad Boettcher
TU Dortmund University
3 shared publications - Sergio Lucia
Technische Universität Dortmund
3 shared publications - Marcel Schade
TU Dortmund University
2 shared publications - Sabrina Grünendahl
TU Dortmund University
2 shared publications B. Pasaliu
Laboratory of Equipment Design, Biochemical and Chemical Engineering, TU Dortmund University, Dortmund, Germany
1 shared publication
Latest Publications
- 19 June 2026
- Published by The CACHE Foundation, Inc.
- Vol. 6, 2457-2465
Enhancing Control in Chemical Processes using Reinforcement from Human Feedback
Reinforcement learning (RL) presents a promising alternative to model-based advanced control schemes, such as model predictive control (MPC), whose application can be limited by highly complex system models. However, incorporating constraints in RL remains challenging and formulating a suitable optimization objective is not straightforward. Reinforcement learning from human feedback (RLHF) offers an approach to derive the RL reward function from human expert preferences, enabling the incorporation of process knowledge. In this work, we present the application of RLHF to fine-tune an approximate MPC controller with suboptimal performance. We demonstrate that combining conventional reward formulations with RLHF, along with varying trajectory segment lengths for collecting human feedback, improves the control methodology for a batch bioreactor by enhancing safety and accounting for long-term effects. Furthermore, direct-preference based policy optimization (DPPO) represents a promising alternative for directly fine-tuning learning-based controllers while circumventing explicit reward model design.
- 19 June 2026
- Published by The CACHE Foundation, Inc.
- Vol. 6, 2457-2465
Enhancing Control in Chemical Processes using Reinforcement from Human Feedback
Reinforcement learning (RL) presents a promising alternative to model-based advanced control schemes, such as model predictive control (MPC), whose application can be limited by highly complex system models. However, incorporating constraints in RL remains challenging and formulating a suitable optimization objective is not straightforward. Reinforcement learning from human feedback (RLHF) offers an approach to derive the RL reward function from human expert preferences, enabling the incorporation of process knowledge. In this work, we present the application of RLHF to fine-tune an approximate MPC controller with suboptimal performance. We demonstrate that combining conventional reward formulations with RLHF, along with varying trajectory segment lengths for collecting human feedback, improves the control methodology for a batch bioreactor by enhancing safety and accounting for long-term effects. Furthermore, direct-preference based policy optimization (DPPO) represents a promising alternative for directly fine-tuning learning-based controllers while circumventing explicit reward model design.
- Open Access
- 18 February 2026
- Published by Elsevier in Computers & Chemical Engineering
- Vol. 209
Computationally efficient Gauss–Newton reinforcement learning for model predictive control
- 18 February 2026
- Published by Elsevier in Computers & Chemical Engineering
- Vol. 209
Open AccessComputationally efficient Gauss–Newton reinforcement learning for model predictive control
- Open Access
- 15 July 2025
- Published by International Association of Online Engineering (IAOE) in International Journal of Online and Biomedical Engineering (iJOE)
- Vol. 21 (09) , 4-26
Redesign of a Virtual Jet Pump Laboratory for Higher Process Engineering Education
The digital transformation increasingly requires the development of competencies for learning and working in the context of Industry 4.0. To meet this need, a practical action research project was conducted to design and implement a virtual instructional laboratory for fluid mechanics in a higher process engineering education course. This paper provides a detailed overview of the key steps and outcomes of the project, including the design of a virtual jet pump experiment using a game engine, the implementation of a real-world scenario (RWS) that was developed with a special constructive alignment (CA) checklist for instructional design of laboratory learning, and the corresponding empirical investigation. The findings demonstrate that the integrated approach employed in this project can enhance learning outcomes related to competencies required in the Industry 4.0 work environment, despite the challenges encountered by students in mastering the RWS. This paper concludes by exploring some potential future directions for further work in the field.
- 15 July 2025
- Published by International Association of Online Engineering (IAOE) in International Journal of Online and Biomedical Engineering (iJOE)
- Vol. 21 (09) , 4-26
Open AccessRedesign of a Virtual Jet Pump Laboratory for Higher Process Engineering Education
The digital transformation increasingly requires the development of competencies for learning and working in the context of Industry 4.0. To meet this need, a practical action research project was conducted to design and implement a virtual instructional laboratory for fluid mechanics in a higher process engineering education course. This paper provides a detailed overview of the key steps and outcomes of the project, including the design of a virtual jet pump experiment using a game engine, the implementation of a real-world scenario (RWS) that was developed with a special constructive alignment (CA) checklist for instructional design of laboratory learning, and the corresponding empirical investigation. The findings demonstrate that the integrated approach employed in this project can enhance learning outcomes related to competencies required in the Industry 4.0 work environment, despite the challenges encountered by students in mastering the RWS. This paper concludes by exploring some potential future directions for further work in the field.
- Top Cited PapersOpen Access
- 7 September 2023
- Published by Elsevier in Control Engineering Practice
- Vol. 140
do-mpc: Towards FAIR nonlinear and robust model predictive control
- 7 September 2023
- Published by Elsevier in Control Engineering Practice
- Vol. 140
Top Cited PapersOpen Accessdo-mpc: Towards FAIR nonlinear and robust model predictive control
- 26 March 2023
- Published by Taylor & Francis in European Journal of Engineering Education
- Vol. 48 (5) , 949-971
Developing a real-world scenario to foster learning and working 4.0 – on using a digital twin of a jet pump experiment in process engineering laboratory education
A newly-developed online laboratory experiment in immersive virtual reality (VR) is used for process engineering students. A digital twin (DT) of a jet pump was realised using the game engine Unreal Engine 4. The chance was taken for an instructional redesign deviating from the conventional cookbook-experiment to better integrate the changing competence demands of the working world 4.0 (W4.0). A real-world scenario (RWS) is used to address the sometimes-vague problem-solving assignments of W4.0: the students found themselves in an ambiguous situation which must be cleared and resolved constructively and autonomously by them. The RWS was evaluated by triangulating qualitative and quantitative methods. Firstly, its Constructive Alignment was evaluated, i.e. whether the intended learning outcomes to be achieved through the novel teaching and learning activities were reflected in the assessment tasks, too. Secondly, students were surveyed regarding their attitudes towards such new forms of laboratory learning. The results show that even if the students initially seem overwhelmed by the unfamiliar type of assignment, they are nevertheless able to solve the work-related engineering problem – at least with varying degrees of assistance from a fictional immediate superior. The students appreciate the opportunity to think creatively and independently develop solutions to the assignment.
- 26 March 2023
- Published by Taylor & Francis in European Journal of Engineering Education
- Vol. 48 (5) , 949-971
Developing a real-world scenario to foster learning and working 4.0 – on using a digital twin of a jet pump experiment in process engineering laboratory education
A newly-developed online laboratory experiment in immersive virtual reality (VR) is used for process engineering students. A digital twin (DT) of a jet pump was realised using the game engine Unreal Engine 4. The chance was taken for an instructional redesign deviating from the conventional cookbook-experiment to better integrate the changing competence demands of the working world 4.0 (W4.0). A real-world scenario (RWS) is used to address the sometimes-vague problem-solving assignments of W4.0: the students found themselves in an ambiguous situation which must be cleared and resolved constructively and autonomously by them. The RWS was evaluated by triangulating qualitative and quantitative methods. Firstly, its Constructive Alignment was evaluated, i.e. whether the intended learning outcomes to be achieved through the novel teaching and learning activities were reflected in the assessment tasks, too. Secondly, students were surveyed regarding their attitudes towards such new forms of laboratory learning. The results show that even if the students initially seem overwhelmed by the unfamiliar type of assignment, they are nevertheless able to solve the work-related engineering problem – at least with varying degrees of assistance from a fictional immediate superior. The students appreciate the opportunity to think creatively and independently develop solutions to the assignment.