Machine Learning Based Methods for Safety and Control of Human–Robot Interaction

Dear Colleagues,

Human–robot Interaction (HRI) research concerns the understanding, design, and evaluation of robotic systems for use by or with humans. The introduction of robots into environments shared with humans could improve the efficiency of both human and robot systems. Using robots to assist humans could increase their precision, speed, and force. Moreover, robots could reduce the stress and tiredness of the human operator, improving working conditions. On the other hand, humans contribute to collaboration in terms of experience, knowledge of how to execute a task, intuition, easy learning and adaptation, and easy understanding of control strategies.

When robots and humans share a workspace, safety is a very important factor due to the operator’s proximity to the robot, which could potentially lead to injuries. Therefore, safety must be considered when designing any HRI control system. Conventional and advanced control methods have been introduced during the last few years, with promising results. The combination of safety with sophisticated controllers based on machine learning approaches will boost the entry of HRI systems in manufacturing, as well as everyday robotic applications.

The existing HRI research is extensive. However, some aspects require special investigation and bridging of the segmented presented research works. Concerning safety methods, a point of interest is the effectiveness of current approaches for detecting collisions (magnitude, direction, position, etc.) with the robot. The human operator can have infinitely different cases of collisions with the robot manipulator; effective identification of these collisions is a crucial point in HRI, which should be thoroughly investigated. This could help to expand current research from robotic manufacturing/factory applications to other robotic sectors, which is a necessity for the robotics community. Furthermore, most of the existing approaches depend on joint torques signals and less on other conventional signals (e.g., joint position or current signals). Hence, there are great HRI systems that could be applied only to collaborative robots, which are more expensive, and less to conventional industrial robots. Concerning control methods, developing controllers for manipulators based on soft computing techniques is required to improve the human–robot co-manipulation. These new approaches should avoid large numbers of computations or those that are complex, and should also avoid expert knowledge for intuitive cooperation. Concerning methods combining both safety and control features, there is a gap where cutting-edge research could be applied. When the implementations of this merging are based on AI and machine-learning algorithms, advanced HRI systems can be achieved. For such systems, effectiveness under extensive different conditions and applicability of the methods to different types of robots and applications could be a key factor of the expected research.

Considering the safety and control of HRI, the focus of this Research Topic is on applying machine learning algorithms, such as supervised learning, unsupervised learning, and reinforcement learning, to the following (and other) topics:

• HRI (physical, social, cognitive).
• Human–robot collaboration (coexistence, synchronization, cooperation, collaboration, co-manipulation).
• Safety in HRI.
• Innovative control approaches.
• Advanced control methods for HRI and HRC.
• Admittance/impedance control.
• Tasks and trajectory planning in HRI.
• Collision detection and reaction.
• Systems for improving HRI performance.
• Measurements in HRI.
• Rehabilitation applications.
• Ergonomics in HRI.
• Human–robot system: productivity, efficiency, and reliability.
• Innovative robotic architectures.
• Safety-related issues.
• Variable stiffness joint.
• Passive/active compliance.

Dr. Abdel-Nasser Sharkawy
Dr. Panagiotis N. Koustoumpardis
Prof. Dr. George Nikolakopoulos
Prof. Dr. Giuseppe Carbone
Dr. Med Amine Laribi
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Machines is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• human–robot interaction
• machine learning
• safety
• collision
• control methods
• collaboration
• coexistence
• cooperative manipulation