Studying how we carry a cup of coffee without spilling it could directly benefit robotics research and development. The way we use tools to get through our daily lives is of great interest to people developing robot manipulators and augmenting devices. In a recent study by Bazzi et al., authors tested the use of contraction theory as a framework for assessing stability and predictability of human movement in complex tasks, like carrying a cup of coffee — an object with internal degrees of freedom.
Their work provides evidence that humans do, in fact, exploit contraction regions, or regions where the trajectory quickly returns to its original path after the system is perturbed, to overcome obstacles during dynamically complex physical interactions.
For these tests, they translated the coffee-transport task into a virtual environment where human subjects were instructed to use a robotic manipulandum to quickly transport a cup containing a ball (a stand-in for coffee) from one end of the screen to the other without losing the ball. Some instances included an applied virtual assistive perturbation, or a force exerted with the direction of the cup, and for others, a virtual resistive perturbance was applied, or a force exerted against the direction of the cup. The authors recorded the force applied by subjects to the manipulandum and the kinematics of the cup and ball.
Results showed that when facing assistive perturbations, the cup trajectories entered a contraction region and stayed inside throughout the duration of the perturbation. Subjects accessed a contraction region after the perturbation in the case of resistive perturbations. When the authors looked at these results as a whole, they concluded that depending on the perturbation type, subjects developed distinctive strategies that exploited the dynamic properties of the task — insightful findings for helping future robotics systems do the same.
Source: “Stability and predictability in human control of complex objects,” by Salah Bazzi, Julia Ebert, Neville Hogan, and Dagmar Sternad, Chaos: An Interdisciplinary Journal of Nonlinear Science (2018). The article can be accessed at https://doi.org/10.1063/1.5042090.