In this article, we develop the method of continuous Pontryagin differentiable programming (Continuous PDP), which enables a robot to learn an objective function from a few sparsely demonstrated keyframes. The keyframes, labeled with some time stamps, are the desired task-space outputs, which a robot is expected to follow sequentially. The time stamps of the keyframes can be different from the time of the robot's actual execution. The method jointly finds an objective function and a time-warping function such that the robot's resulting trajectory sequentially follows the keyframes with minimal discrepancy loss. The Continuous PDP minimizes the discrepancy loss using projected gradient descent by efficiently solving the gradient of the robot trajectory with respect to the unknown parameters. The method is first evaluated on a simulated robot arm and then applied to a 6-DoF quadrotor to learn an objective function for motion planning in unmodeled environments. The results show the efficiency of the method, its ability to handle time misalignment between keyframes and robot execution, and the generalization of objective learning into unseen motion conditions.

This paper presents an empirical study of two student pairs collaborating on two small products design session in both face-to-face and distributed settings while using computer-mediated communication (CMC) technologies and a collaborative virtual environment (CVE). To gain insight about the way designers communicate and collaborate, the observation focused on how much time the students worked together and individually in the design process. The study shows that teams worked together more in the distributed setting than in the face-to-face setting. In the post-test questionnaires, participants reported that they found the distributed setting a more engaging environment to work with teammates. Findings of the study suggest plausible design criteria for a communication system for distributed collaboration that supports interaction and sharing design information.