Rich Physical Human/Robot Interaction Skills in Versatile Real/Life Scenarios

Principal Investigator:

TRI Liaison:

Sam Zapolsky

Project Summary

Develop a collaborative sensor-based semiautonomous control framework where:
- the robot is in charge of performing the physically demanding/mundane tasks
- the human provides direct physical collaboration, remote teleoperation, or high-level cognitive guidance.
Requires: close control loop with multi-leveled perception.
Implement and verify on a multi-armed mobile platform under heavy and complex physical contact with the environment

Research Goals

Develop multi-armed haptic teleoperation – controlling robot task points through force-feedback interfaces

Design virtual fixtures
Design control primitives based on human skills
Implement transitions between control primitives
Design a task observer estimating contact situations, detecting execution errors, and verifying task completion
Develop control fusion framework for human-robot collaboration