UC robotics research set to transform NZ horticulture industry

Published by Communications and Development

12 July 2010

A vine-pruning robot that could save New Zealand’s horticulture industry $27.5 million a year is being developed by a University of Canterbury-led research team.

The team led by Dr Richard Green (Computer Science and Software Engineering) has received almost $3 million in funding from the Foundation for Research, Science and Technology for their programme to develop an intelligent vision-based pruning system.

Vine-pruning robot researchers

(From left) Dr Richard Green (Computer Science and Software Engineering), John Duncan and Heather Thomas (Research and Innovation), Dr Val Saxton from Lincoln University, Dr Dean Kirk (Product Innovation Centre), and Professor XiaoQi Chen (Mechanical Engineering).

The world leading research team includes UC’s Professor Tanja Mitrovic (Computer Science HOD), Professor XiaoQi Chen (Mechatronics Director), Dr David Aitchison (Mechanical Engineering), Dr Dean Kirk (Product Innovation Centre) and Lincoln University viticulture researcher Dr Val Saxton. The team plan to spend the next four years in the research and development of advanced vision-based real time 3D modelling, interfaced to multiple high performance robotic arms with cutters to accomplish economically automated pruning.

The robotic system, which will be manufactured in New Zealand, is forecast to earn New Zealand exporters over $200 million within 10 years of market entry. It is also estimated to provide savings of $27.5 million per annum to the New Zealand wine industry through increased productivity and reduced yield losses. 

“Such a fast vision-based pruning system is only possible using recently developed camera technology with efficient cutting edge computer vision-based tracking and AI algorithms,” said Dr Green. “We are leading the world with fast accurate colour 3D depth maps of vines, light robot cutting arms and the AI to coordinate this moving at walking speed. Not only can a higher quality be maintained by pruning consistently and accurately while recognising disease and age of vines, but the industry will be able to guarantee pruning within the very brief seasonal window each year.”

The robotic technology will use artificial intelligence to recognise plant features and synchronise multiple cameras and high-speed robot arm pruners with immediate application to vine pruning and a longer-term broader application to general harvesting and pruning in the agriculture industry.

“This is not just an excuse to combine drinking wine with research, but our internationally renowned interdisciplinary research team is looking forward to confirming the quality of the results of the vine-pruning robot,” Dr Green said.

The research is being undertaken with the support of Scott Technology Ltd, developers of image-based robot systems, the New Zealand Wine Growers Association and the largest NZ wine and spirit company Pernod Ricard NZ Ltd.

For further information please contact:

Dr Richard Green
Senior Lecturer
Computer Science and Software Engineering department
University of Canterbury
Ph: +64 3 364 2987 ext 8436
Mobile: +64 21 331 707