Human-like Robotic Controllers for Enhanced Motor Learning

Summary

The purpose of this study is to develop a new paradigm to understand how humans physically interact with each other at a single and at multiple joints, with multiple contact points, so as to synthesize robot controllers that can exhibit human-like behavior when interacting with humans (e.g., exoskeleton) or other co-robots. The investigators will develop models for a single joint robot (i.e. at the ankle joint) that can vary its haptic behavioral interactions at variable impedances, and replicate in a multi-joint robot (i.e. at the ankle, knee, and hip joints). The investigators will collect data from healthy participants and clinical populations to create a controller based on our models to implement in the robots. Then, the investigators will test our models via the robots to investigate the mechanisms underlying enhanced motor learning during different human-human haptic interaction behaviors (i.e. collaboration, competition, and cooperation. This study will be carried out in healthy participants, participants post-stroke, and participants with spinal cord injury (SCI).

Description

Eligibility

Age range

18–80 years

Sex

All

Healthy volunteers

Yes

Interventions

• Behavioral
  Behavioral Interaction Conditions

  The participants will be single-blinded and complete a tracking task as either: solo task, collaboration task (both participants work on a common task synchronously to achieve a goal; this is a summative effort to achieve the goal), competition task (each participant has to achieve a goal at the expense of his or her partner, therefore maximizing effort or error of the partner in reaching the goal), or cooperation task (an asymmetric partnership with an active partner and a passive partner working towards a goal).

• Device
  Haptic Impedance Level

  The subjects will complete their task at 3 impedance levels: high (a virtual stiffness 160-200 N/m and damping 0\~10 Nm/s; this will be a stiff connection in which the subjects feel like they are connected via rigid links and each subject will perceive the other partner's movement directly), medium (a virtual stiffness 100-140 N/m and damping 0\~10 Nm/s; this will be a spring like-connection in which the subjects feel like they are connected with a spring and each subject will perceive the other partner with a force that is proportional to the trajectory difference of the two participants), and soft (a virtual stiffness 40-80 N/m and damping 0\~10 Nm/s; this will be a spring like connection in which the subjects feel like they are connected with a loose spring and each subject will perceive the other partner with a force that is proportional to the trajectory difference of the two subjects, however, this force will be smaller than that of the medium impedance).

• Behavioral
  Skill Level of Partner

  There will be two skill levels: novice (a participant who has no prior experience with the trajectory tracking experiment; in testing with the clinical populations, the investigators will assign this condition to the clinical participant) and expert (a participant who is experienced with the trajectory tracking experiment and who can achieve a tracking error \[difference of the desired trajectory and actual trajectory\] below a certain threshold; in testing with the clinical population, the investigators will assign this condition to the therapist). Participants will start experimentation paired as novice-novice, and at the end of the session may be invited to continue additional sessions to be paired as the expert in a novice-expert dyad.

Location