3rd Workshop on Robots in Clutter: Perception and Interaction in Clutter

IEEE/RSJ International Conference on Intelligent Robots and Systems
September 18th, 2014, Chicago, Illinois

Complex and cluttered environments continue to present challenging problems to many aspects of robotics research. Vision for example faces the problem of segmenting or recognising task-relevant objects amidst clutter and occlusions. Unexpected scene changes pose challenges for maintaining valid and tractable scene representations for navigation, especially in highly dynamic outdoor scenes as encountered in self-driving cars. Manipulation cannot expect precise pose knowledge of all objects in a pile, let alone all contact relations. All these problems will become increasingly manifest as robots move into unstructured domestic, industrial or outdoor settings. What is meant by "robust to clutter", however, is difficult to define and adequate metrics and benchmarks are still missing.

Humans seem to be good at handling these situations by identifying the most task relevant parts of the scene and focusing attention on these, or by employing interactive strategies where a problem is tackled in a series of small steps. Similar strategies are also being increasingly explored in various areas of robotics.

Following in the footsteps of the first and second Workshop on Robots in Clutter held at RSS 2012 and 2013 respectively, this workshop brings together researchers from related robotics domains to discuss experience and ideas for handling various problems induced by clutter, and to advance theoretically founded and system-wide approaches of handling clutter.

To establish problems and methods related to clutter as a distinct and recognisable topic within the community we will pursue the publication of selected workshop contributions as a special journal issue. We will furthermore use the plenary discussion at the workshop to form a working group which should create a roadmap in order to systematically drive this topic further.

This workshop is supported by the IEEE RAS Technical Committee on Mobile Manipulation (TCMM) and by EU project FP7-ICT-610532 SQUIRREL: Clearing Clutter Bit by Bit".