It’s perhaps easy to fall into the trap of regarding the use of robotics as an especially industrial endeavor. Among the most enthusiastic adopters of robotic technology has been agriculture however, with previous posts exploring the use of autonomous technology to do everything from monitor soil health to harvest crops.
A good example of the work being done in the field comes via the EU’s CATCH project, for which the Fraunhofer Institute for Production Systems and Design is working on robotic technology that can autonomously harvest cucumbers.
In a pickle
Cucumbers that are destined for pickle jars are traditionally harvested by hand, with the only technological assistance coming via ‘cucumber flyers’. It’s particularly back breaking work, with pickers required to lie on their stomachs on board the wings of the flyer and pluck out every ripe cucumber.
The CATCH project aims to develop a dual-arm robotic system that can not only automate cucumber farming but also various other agricultural procedures. The robot has been equipped with tactile perception technology, whilst it’s also capable of performing in a range of conditions. The idea is to replicate many of the motions performed by human pickers, and therefore pluck the cucumbers without damaging them. What’s more, it also needs to be able to match the speed of human pickers, and pluck at least 13 cucumbers per minute.
Doing this successfully represents a considerable challenge. The cucumbers themselves are challenging to handle, whilst the environment they grow in can make it difficult to identify them successfully. For instance, they’re usually randomly distributed throughout the field, whilst varying light conditions can make spotting them tough.
The team hope to achieve this via the use of multispectral cameras and smart image processing. The system developed to date has been able to accurately detect and locate cucumbers with 95% accuracy.
Put to the test
The system was put through its paces in the summer of 2017 in field tests conducted in Germany at the Leibniz Institute for Agricultural Engineering and Bioeconomy. It was able to harvest various types of cucumber successfully in both the field and in greenhouse environments.
The team next plan to examine the impact potential malfunctions in the system might have on the accuracy of the picking, before then moving onto commercial rollout of the system. They’ve already had interest in it from cucumber farmers, agricultural associations and other companies.
“Our goal is to create an intelligent control system capable of making judgment calls: assigning a certain task to a certain gripper arm, monitoring cucumber picking and dealing with exceptions,” the team say.