Grant: Discovery Grants Program - Individual

Robot manipulators are expected to execute manipulation tasks of increased complexity in unstructured human and space environments, which calls for advancement in robot control techniques and robot intelligence. New research challenges emerge as robots are increasingly desirable to interact, assist, serve, and explore with humans. In tackling these challenges, intelligent robot manipulators with multiple working modes are essential. As robots continue to complement humans in their work, safety has been a critical concern in robot design in terms of actuation, sensing and control. In line with the most recent research advancements on robotic manipulation, the long-term vision of the proposed research program is to develop intelligent robot systems for efficient and safe robotic manipulations in unstructured human and space environments. This will involve the development of innovative robot manipulators with capable intelligent control systems. Aiming at this long term objective, and building upon our recent research achievements, the following short-term objectives are defined: 1) develop a framework for robot manipulation control by generalizing our proposed multiple working mode (MWM) robot control approach, categorizing and accommodating model inaccuracies associated with unstructured environments; 2) develop robot intelligence by fusing the proposed MWM robot control approach with machine learning techniques for complicated robotic manipulation tasks with environment model inaccuracy and real-time variation; 3) extend the MWM control approach to robotic capture of unknown targets in space, applying the recently developed methods for joint torque estimation and tip position measurement; 4) further develop our spring assisted modular and reconfigurable robot and control system for dexterous and safe manipulations with multiple working modes and robot intelligence, with consideration of practical issues such as energy efficiency and emergency braking with a spring and clutch mechanism; and 5) develop software for innovative programming of intelligent robot manipulators with multiple working modes for complicated manipulation tasks. The proposed research program will generate innovative yet practical techniques of design and control of intelligent robotic systems for performing complex manipulations in unstructured human or space environments, generating high quality research contributions and providing effective solutions to practical robotic manipulation problems. Robot systems capable of manipulating effectively in unstructured environments will find tremendous applications in sectors such as service, health, advanced manufacturing and space exploration, leading to huge commercialization potentials. Furthermore, with its multidisciplinary nature, the proposed research program will contribute significantly to training of highly qualified personnel for Canada.