The editorial emphasizes that the huge variety of interaction scenarios, as well as the type of interaction, leads to a similarly large spectrum of models and methods for designing machines capable of completing tasks together with a human. The articles show different ways in which the interaction between humans and machines can take place, from shared steering of automated vehicles to haptic joysticks for wheelchairs and cooperative path following for underwater robots. One of the most prominent application domains of cooperative and shared control systems is highly automated driving, and this is reflected in two articles reporting on research in that field.

The first, titled A review of shared control in automated vehicles: System evaluation by Sarabia and colleagues, acknowledges the pivotal role of human-machine cooperative driving in future transportation and provides a survey of recent literature focusing on studies with real drivers. The authors emphasize the importance of turning attention to the evaluation methodology of shared control approaches, reviewing both study design and the various metrics used in the studies. Among their conclusions, the editorial notes that research on shared longitudinal control should catch up with research on shared lateral control, that control authority shifting based on driver state should keep receiving increasing attention, and that the evaluation of the interaction itself and of safety have not been performed as intensively as the evaluation of driving performance.

The second automotive contribution, by Oudainia and colleagues, aims at minimizing conflicts between drivers and lane keeping systems of intelligent vehicles. In their article, Online driver model parameter identification using the Lyapunov approach based shared control, the authors present a driving assistance system for lane keeping that is based on a dynamic driver model. To robustly identify the driver model parameters for different drivers and situations, a Lyapunov approach is applied, and using this approach the asymptotic stability of the closed-loop control system can be shown. The resulting control design was applied to a driving simulator, where experimental results demonstrated the effectiveness of the new approach and the utilization of the identified parameters in a shared control design. In addition to an objective performance increase in terms of improved safety and reduced conflicts, the drivers’ subjective perception of the assistance system improved as well.

The scopes of the other articles are closely related to the automotive application domain and focus on cooperative human-machine system design for mobile nursing robots that accompany patients, the control of power wheelchairs, and tele-operated underwater vehicles. In the article Negotiation-based cooperative planning of local trajectories, Schneider and colleagues consider a situation in a hospital where a nursing robot has to accompany patients who need support for walking from one room to another. One of the key aspects in this scenario is the dynamic negotiation of roles that permits both the human and machine to take the initiative in order to follow a particular path. The authors use a negotiation theoretical approach to model the interaction within the human-machine system and propose a reciprocal tit-for-tat strategy to achieve an agreement concerning the trajectory plan, without resorting to an imposed time-based deadline. A simulation study shows how the results of the cooperative trajectory planning change with different kinds of human behavior.

In the article Model-based shared control approach for a power wheelchair driving assistance system using a force-feedback joystick, Nguyen and colleagues present a haptic-based assistance system for power wheelchair users to avoid collisions and reduce user effort. Users operate the wheelchair via a joystick capable of haptic feedback, and the applied shared-control strategy is composed of an operational part and a tactical part. On an operational level, a Takagi-Sugeno fuzzy control approach addresses the time-varying user intentions when supporting the user’s driving actions. On a tactical level, conflicts between the assistance system and the user are managed based on the driver’s hand torque estimation, ensuring that the user has the final decision. The assistance system proved its effectiveness in an experimental evaluation.

Another interesting area of application for human-machine cooperative systems is investigated in the work of Sato and colleagues, titled Cooperative path-following control of a remotely operated underwater vehicle for human visual inspection task. In the context of remotely operated underwater vehicles used for inspecting underwater infrastructure, the authors developed an assistance system to reduce the workload of human operators. The assistance system ensures that the underwater vehicle robustly follows a predefined path while the human operator can focus on determining the vehicle’s velocity to properly conduct the inspection task. To validate the system, the article presents an experiment comparing system performance with and without the assistance system.

The results of the underwater vehicle study show the advantages of applying the assistance system compared to manual control in terms of significantly reduced path-following errors and operator workload. Moreover, the cooperative path-following control method enables the intended shift of operator attention, which ranges from tight control of the robot’s motion to the inspection task itself. This division of labor illustrates a central theme of the Research Topic: systematic, model-based approaches can lead to an effective human-machine system in which the strengths of both partners are exploited.

Overall, the five contributions to this Research Topic provide a contemporary insight into research on human-machine cooperative and shared control systems from different application domains. The articles demonstrate the multifaceted nature of human-machine cooperation and show how the use of systematic approaches can lead to an effective human-machine system. The editorial is presented to encourage research to continue in this direction and to spread to other fields of application, from advanced driver assistance systems and powered wheelchairs to cooperative planning and remotely operated vehicles.

The editorial, received on 17 March 2023, accepted on 27 March 2023, and published on 11 April 2023 in Volume 4 of Frontiers in Control Engineering, was edited and reviewed by Jie Chen of City University of Hong Kong. It was submitted to the Networked Control section of the journal, and the Research Topic it introduces covers keywords including shared control, human-robot interaction, human-machine cooperation, powered wheelchair, cooperative planning, remotely operated vehicle, and advanced driver assistance systems. As the field matures, the editorial suggests that the evaluation of interaction quality and safety deserves as much attention as driving performance, and that model-based methods will remain central to designing machines that can genuinely cooperate with humans rather than merely replace them.

Subject of Research: Editorial: Model-based methods for human–machine cooperative and shared control systems

Article Title: Editorial: Model-based methods for human–machine cooperative and shared control systems

Article References: Inga, J., Rothfuß, S., Saito, Y., & Sentouh, C. (2023). Editorial: Model-based methods for human–machine cooperative and shared control systems. Frontiers in Control Engineering, 4, Article 1188846. https://doi.org/10.3389/fcteg.2023.1188846

Image Credits: AI Generated

DOI: 10.3389/fcteg.2023.1188846

Keywords: Editorial, Model-based, methods, human, machine, cooperative, shared, control, systems, scientific research