The aim of the workshop is to discuss current problems and future directions in contact motion generation. Recent years have seen significant progress toward enabling autonomous mobile robots to interact with unstructured, dynamic, and human-centered environments. Much of this progress, however, is concerned with collision-free motion. For example, most popular planning algorithms utilize simplistic, binary models of contact. Methods based on distance fields can reason about the contact in a continuous manner, but remarkably few approaches leverage this capacity in a deliberate, systematic fashion. In contrast, the goal of mobile manipulation is to interact with objects, to create contact and to exploit it judiciously in order to change the state of the world.

In mobile manipulation, motion generation must also consider a larger state space and more pronounced effects of uncertainty than in static manipulation problems. To leverage progress in motion generation, therefore, a closer exchange between researchers in control, motion planning, and mobile manipulation is required. A workshop on this topic would enable a focused discussion on the key issues that prevent the field from better exploiting contact, including improved models of contact, related sampling quality issues and algorithmic approaches that lend themselves well to contact motion generation.