Reaching is Better When You Get What You Want: Realtime Feedback of Intended Reaching Trajectory Despite an Unstable Environment

Justin Horowitz, Tejas Madhavan, Christine Massie, James Patton

Improvements in human-machine interaction may help overcome the unstable and uncertain environments that cause problems in everyday living. Here we experimentally evaluated intent feedback (IF), which estimates and displays the human operator's underlying intended trajectory in real-time. IF is a filter that combines a model of the arm with position and force data to determine the intended position. Subjects performed targeted reaching motions while seeing either their actual hand position or their estimated intent as a cursor while they experienced white noise forces rendered by a robotic handle. We found significantly better reaching performance during force exposure using the estimated intent. Additionally, in a second set of subjects with a reduced modeled stiffness, IF reduced estimated arm stiffness to about half that without IF, indicating a more relaxed state of operation. While visual distortions typically degrade performance and require an adaptation period to overcome, this particular distortion immediately enhanced performance. In the future, this method could provide novel insights into the nature of control. IF might also be applied in driving and piloting applications to best follow a person's desire in unpredictable or turbulent conditions.