Le 30/05/2022

Ekobot: New autonomous weeding experience

A new machine is put at work on farms in Sweden and The Netherlands. It is weeding with three core components: By cameras (1) that are looking at the crop. The essential intelligence (2) that’s making the decisions. And an arm that will give the actual strike (3).

The Ekobot robot is making use of existing electrical driving platform. Navigation is done with a combination of optical guidance and RTK GPS systems.

Field of onions

The weeding can be done in and between the rows of onions. On the field it’s looking at each onion plant, in order to create a ‘safety margin’ around the plant. Other items that are detected as weed, will get a mechanical strike if they are outside the safety zone. It is able to operate with a speed up to 1 km per hour. With thousands and thousands of strikes per day.

Based on working with the robot on the field, one can see and count that hardly any onions are damaged. Meanwhile, the large majority of weed is being removed.

The weed when it’s just started to grow, is simply the best to strike. The Ekobot robot will flip it up-side down with a strike. The sun and wind will take further care of ‘drying-out’ the roots of the weed plant. In order to be on time, a frequent pass in the field will be the solution.

Daily commute

Ekobot weeding robot - @ducksize

It’s easy to imagen this can work on a daily routine basis, by the push of a button (or not even..). Once everything is set-up properly, the initiation of the robot is a flip of switch. Off course, there is the need to learn the robot the field. By showing where the rows are. The turning at headlands will be calculated automatically. In the current set-up, charging is needed after each run on a day.

More hands-on experience and further information

In the field, Corné Rispens from ducksize is hands-on active in the team to operate and introduce these robots. In the upcoming weeks, more experience will be shared at Ekobot robot field experience page at ducksize.com

Pictures credit @ducksize

Categories : #Robots
  • Corné Rispens
    Founder at ducksize.com