Authors:
Sune Sørensen
1
;
Lakshadeep Naik
2
;
Peter Nguyen
2
;
Aljaz Kramberger
2
;
Leon Bodenhagen
2
;
Mikkel Baun Kjærgaard
1
and
Norbert Krüger
2
;
3
Affiliations:
1
SDU Software Engineering, The Mærsk McKinney Møller Institute, University of Southern Denmark, Campusvej 55, Odense, Denmark
;
2
SDU Robotics, The Mærsk McKinney Møller Institute, University of Southern Denmark, Campusvej 55, Odense, Denmark
;
3
Danish Institute for Advanced Studies (DIAS), Odense M, Denmark
Keyword(s):
Base Pose Planning, Mobile Manipulation, Dynamic Programming, World Modeling, Perceptual Anchoring.
Abstract:
Given a setup with external cameras and a mobile manipulator with an eye-in-hand camera, we address the problem of computing a sequence of base poses and grasp choices that allows for clearing objects from a table while minimizing the overall execution time. The first step in our approach is to construct a world model, which is generated by an anchoring process, using information from the external cameras. Next, we developed a planning module which – based on the contents of the world model - is able to create a plausible plan for reaching base positions and suitable grasp choices keeping execution time minimal. Comparing our approach to two baseline methods shows that the average execution cost of plans computed by our approach is 40% lower than the naive baseline and 33% lower than the heuristic-based baseline. Furthermore, we integrate our approach in a demonstrator, undertaking the full complexity of the problem.