We propose a new policy class, Composable Interaction Primitives (CIPs), specialized for learning sustained-contact manipulation skills like opening a drawer, pulling a lever, turning a wheel, or shifting gears. CIPs have two primary design goals: to minimize what must be learned by exploiting structure present in the world and the robot, and to support sequential composition by construction, so that learned skills can be used by a task-level planner. Using an ablation experiment in four simulated manipulation tasks, we show that the structure included in CIPs substantially improves the efficiency of motor skill learning. We then show that CIPs can be used for plan execution in a zero-shot fashion by sequencing learned skills. We validate our approach on real robot hardware by learning and sequencing two manipulation skills.

We propose a new, data-efficient approach for skill transfer to novel objects, accounting for known categorical shape variation. A low-dimensional shape representation embedding is learned from a set of deformations, sampled between known objects within a category. This latent representation is mapped to a set of control parameters that result in successful execution of a category-level skill on that object. This method generalizes a learned manipulation policy to unseen objects with few training examples. We demonstrate this approach on pouring from cups and scooping with spatulas, where there is complex, nonlinear variation of successful control parameters across objects.

Distributed manipulators - consisting of a set of actuators or robots working cooperatively to achieve a manipulation task - are robust and flexible tools for performing a range of planar manipulation skills. One novel example is the delta array, a distributed manipulator composed of a grid of delta robots, capable of performing dexterous manipulation tasks using strategies incorporating both dynamic and static contact. Hand-designing effective distributed control policies for such a manipulator can be complex and time consuming, given the high-dimensional action space and unfamiliar system dynamics. In this paper, we examine the principles guiding development and control of such a delta array for a planar translation task. We explore policy learning as a robust cooperative control approach, allowing for smooth manipulation of a range of objects, showing improved accuracy and efficiency over baseline human-designed policies.