Autonomous robotic algorithms enhance mobility for people with motor impairments. We develop ramp traversal and drop-off detection systems for powered wheelchairs to improve safety and accessibility.
Robotic autonomy can be leveraged to enhance human motor learning, improving accessibility to advanced assistive technologies. We explore the potential of Body Machine Interfaces to control high dimensional robotic arms.
Robot Learning from Motor-Impaired Teachers
Standard machine learning relies on expert data. We address the challenge of learning directly from motor-impaired users, who may lack task expertise and face limitations in providing demonstrations, yet are key to designing effective assistive robots.
Interface-Aware Robotic Intelligence
Interface-awareness explicitly represents uncertainty in human commands. We model how physical interface actions map to robot actions & how the interface alters control signals to aid autonomy in identifying and improving teleoperation deficiencies.
Shared control leverages robotics and human strengths to enhance safety, efficiency, and task success. We analyze human signals and develop algorithms to enhance human-robot shared control for diverse users, specifically those with motor impairments.
We explore soft and wearable sensors, from dielectric elastomers for sensing and energy generation to HRV and IMU sensors that measure workload and harness residual motion, enabling better control of assistive devices and improving accessibility.
Accessibility of Robotics Research
Robotics has the potential to make the world more inclusive. We address barriers for motor-impaired individuals in robotics, advocating for their representation as researchers and tackling challenges to foster inclusion.