Download coming soon...
An approach for adaptive shared control of an assistive manipulator is presented. A set of distributed collision and proximity sensors is used to aid in limiting collisions during direct control by the disabled user. Artificial neural networks adapt the use of the proximity sensors online, which limits movements in the direction of an obstacle before a collision occurs. The system learns by associating the different proximity sensors to the collision sensors where collisions are detected. This enables the user and the robot to adapt simultaneously and in real-time, with the objective of converging on a usage of the proximity sensors that increases performance for a given user, robot implementation and task-set. The system was tested in a controlled setting with a simulated 5 DOF assistive manipulator and showed promising reductions in the mean time on simplified manipulation tasks. It extends earlier work by showing that the approach can be applied to full multi-link manipulators.
@Inproceedings{Stoelen2013,
author={Stoelen, M.F. and Tejada, V.F. and Victores, J.G. and Jardoan Huete, A. and Bonsignorio, F. and Balaguer, C.},
booktitle={Intelligent Robots and Systems (IROS), 2013 IEEE/RSJ International Conference on},
title={Adaptive collision-limitation behavior for an assistive manipulator},
year={2013},
pages={1143-1148},
doi={10.1109/IROS.2013.6696494},
ISSN={2153-0858}}
All joint and Cartesian trajectories. All trial-level results. All proximity sensor readings over time. All shared control output over time. Development of neural network weight matrices over time. Matlab example for loading the trajectories of all successful attempts. Pre and post questionnaires, consent form with participant instructions. See [AdaptiveCollisionManipulator-DataSets].
Complete set of executables, Linux 32 and 64 bit. Tested on Ubuntu 10.04 for 32 bit and 11.04 for 64 bit. See [AdaptiveCollisionManipulator-Code].
Simulated robot only. Multi-core desktop computer, 8-core Intel i7-2600 @ 3.4 GHz used here. 3DConnexion SpaceNavigator 6 DOF input device. 3D capable display, 40 inch (approx. 102 cm) Samsung 3D TV (UE40D8000) with active 3D glasses used here. See [AdaptiveCollisionManipulator-Hardware].
