Simulation, Learning and Control Methods to Improve Robotic Vegetable Harvesting
Accepted version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Birrell, Simon https://orcid.org/0000-0002-3636-0177
Iida, Fumiya https://orcid.org/0000-0001-9246-7190
Abstract
Agricultural robots are subject to a much harsher envi- ronment than those in the factory or lab and control strategies need to take this into account while maintaining a low cycle time. Three control strategies were tested on Vegebot, a lettuce-picking robot, in both simulation and on the real robot. Between a fast open loop that was vulnerable to environmental noise and a slow but robust visual servoing technique, a Learned Open Loop strategy was tested where the robot learned from successful picks to pick at an intermediate speed. This reduced the projected cycle time from 31s to 17.2s, a 45% reduction.
Description
Keywords
40 Engineering, 46 Information and Computing Sciences, 4007 Control Engineering, Mechatronics and Robotics, 4602 Artificial Intelligence, 4010 Engineering Practice and Education
Journal Title
Springer Proceedings in Advanced Robotics
Conference Name
17th International Symposium on Experimental Robotics
Journal ISSN
2511-1256
2511-1264
2511-1264
Volume Title
19
Publisher
Springer International Publishing
Publisher DOI
Rights
All rights reserved
Sponsorship
Royal Society (TA160113)
Engineering and Physical Sciences Research Council (EP/L015889/1)
Engineering and Physical Sciences Research Council (EP/L015889/1)
This project was possible thanks to EPSRC Grant EP/L015889/1, the Royal Society ERA Foundation Translation Award (TA160113), EPSRC Doctoral Training Program ICASE AwardRG84492 (cofunded by G’s Growers), EPSRC Small Partnership AwardRG86264 (in collaboration with G’s Growers), and the BBSRC Small Partnership GrantRG81275. Special thanks to G Growers, George Walker and Josie Hughes for their invaluable assistance.