Funded by an EPSRC Fellowship, ‘FREEpHRI: Flexible, Robust and Efficient Physical Human-Robot Interaction with iterative learning and self-triggered role adaption’ aims to allow robots to intelligently detect the changes of human behaviours and automatically adjust the human-robot relationship.
“In recent decades, robots have demonstrated their superiority in increasing productivity, efficiency and consistency in the manufacturing of products,” said principal investigator Dr. Zhenhong Li, a research fellow in Leeds’ School of Electronic and Electrical Engineering.
“Now, we aim to put humans back into the development and production process. By taking the advantage of robots’ precision and humans’ creativity, we may create a better system that can provide individualised products and services.”
The fellowship will focus on the development of a human-robot interaction control strategy which enables robots to flexibly respond to their human partners, Li told The Engineer, for intended applications across industries such as manufacturing and healthcare.
“Leveraging the latest development of optimisation theory and machine learning algorithms, we are going to investigate three fundamental issues in physical human-robot interaction,” Li said.
These, he explained, are how to model complex human interaction behaviours; how to efficiently update the robot’s control strategy to ensure desired interactions; and how to deal with uncertainties in the human-robot system.
According to the team’s EPSRC grant summary, the control strategy will use optimisation theory to model human-robot interaction behaviours and learning techniques to compensate the effects of unknown dynamics and external disturbances.
The human partner will be assigned a cost function implying motor capability, allowing the robot to adjust its role (collaborator or competitor) according to the real-time estimation of the human cost function.
A self-triggered role adaption mechanism will use the performance of the human-robot system and estimated human behaviour to detect the role changes of the human, triggering the robot to change its role when necessary.
Reliability and functionality of the proposed techniques will be evaluated through application in physical robot-assisted rehabilitation, according to the team, with the techniques being used to achieve typical training strategies initially in lab settings, then in hospital settings supported by the Leeds Teaching Hospital.
A strategic group will be established to inform design and test of the technologies, drawing on the expertise of academic and industry partners.
Poll finds engineers are Britain’s second most trusted profession
Interesting. Government ministers are nearly 50% more trusted than politicians! "politicians (11 per cent ), government ministers (15 per...