Sign up for our daily Newsletter and stay up to date with all the latest news!

Subscribe I am already a subscriber

You are using software which is blocking our advertisements (adblocker).

As we provide the news for free, we are relying on revenues from our banners. So please disable your adblocker and reload the page to continue using this site.

Click here for a guide on disabling your adblocker.

Sign up for our daily Newsletter and stay up to date with all the latest news!

Subscribe I am already a subscriber

Machine locates fruits and measures level of ripeness

Remember the video earlier this year, in which a robot drove through the greenhouse, selected ripe tomatoes and harvested them?  Remember questioning whether it will be a real thing in the future - a machine that can detect ripe fruits, measures their location and harvests them right away? Well, we've spoken with MetoMotion, developer of the GRoW robot, and got an update for you. 

In a new video Metomotion shows their system establishing the ripeness of the fruits. "Indeed the algorithm in the video is the next version of our vision system", Adi Nir with MetoMotion explains. "The vision system is a 3D stereo vision system that detects the ripe fruits and measures their location in space. The algorithm is based on thousands of pictures taken from different locations."  

By using deep learning technology the system was trained to locate the fruits and measure their level of ripeness. "It detects accurately the fruits ready for harvest by looking at all the fruits in the vine. The decision whether to pick the vine based on the ripeness level of all the fruits in the vine giving ripeness grade for each individual fruit."

The grower has the flexibility to adjust the robot when to pick the tomatoes according to the different varieties and his professional needs. "In addition, all the data collected and analyzed to give the grower information about the crop like yield quantity map of his greenhouse, and a forecast of the yield", says Adi. The data is also used for optimal planning of picking schedule for the robot.

The algorithm will be implemented in GRoW, MetoMotions tomato harvesting robot. After a successful alpha pilot in Israel earlier this year, MetoMotion continues the development of the tomato harvesting robot. "We're working on the beta version with improved performance. The new version will have an improved vision system, an updated harvesting tool, the robot will work faster and at higher quality." 

The company plans to start trials with Beta version of GRoW in a few months in the Netherlands. GRoW tomato harvesting robot is expected to be commercially available on 2020. "We are looking to cooperate with business partners and growers in this." 

For more information:
Tel: +972-72-260-7000 

Publication date: