Physical cooperation between a person and a robot requires a range of performance characteristics that is difficult for traditional robots to meet. Sometimes a robot must have high stiffness to support an object for a person to work with or to perform a precise task of its own, but at other times it must have low stiffness to ensure the safety of a person who makes contact with it. Sometimes a robot must control the interaction force between itself and a person, but at other times it must move along a precise path regardless of the force applied to it. Existing actuators such as electric, pneumatic or hydraulic motors cannot achieve this performance combination, so a new approach to robotic actuators is required. This National Robotics Initiative (NRI) research project will study the design and control of robots that use a combination of active components (motors) and passive components (brakes) to meet these opposing demands. This actuation approach, referred to as Balanced Active-Passive Hybrid Actuation, will provide high power capabilities while possessing the unique characteristics required for human-robot physical collaboration. This research will enable new capabilities for cooperation between people and robots and will open unserved application areas such as cooperative high-power manufacturing robots, high-power rehabilitation robots, and high-performance exoskeletons. One important application is rehabilitation robotics for retraining intentional movement after neural injury such as stroke. The hybrid active-passive actuators developed in this research will be applied to create a strong and accurate, yet also safe, rehabilitation robot for the legs, which will be used to test leg force and movement control and develop rehabilitation methods for stroke survivors.

The goal of this project is to develop design principles and control approaches for hybrid active-passive actuators capable of providing high power and high force-control bandwidth. The research will combine a series-elastic electric motor for low-bandwidth force control, a parallel direct-drive electric motor for high-bandwidth force control, and a parallel brake for efficient very high-bandwidth support against applied external forces. A physical test actuator will be built according to these design principles and used to develop and test a bandwidth-partitioning controller for coordinating the multiple actuators. The resulting design and control techniques will be used to build a two-axis parallel linkage manipulator for applying arbitrary planar force fields to the foot of a human user. The system will be tested for its ability to haptically render virtual mechanical environments such as viscous curl fields. These curl fields will be used to study motor control of the leg, with the goal of developing rehabilitative tasks for stroke survivors to promote the recovery of coordinated movement and postural control.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Project Start
Project End
Budget Start
2018-09-15
Budget End
2021-08-31
Support Year
Fiscal Year
2018
Total Cost
$749,704
Indirect Cost
Name
University of Wisconsin Madison
Department
Type
DUNS #
City
Madison
State
WI
Country
United States
Zip Code
53715