Intelligent Hydraulic Systems in Agricultural Machinery

Intelligent Hydraulic Systems in Agricultural Machinery

Weeds can be a big problem for farmers. Is there an alternative to widespread spraying?

Our Deepfield Robotics team has been working on a solution. We’re currently developing prototype four-wheeled machines. They’re compact, and they work in groups, moving up and down rows of crops. Each of them is fitted with cameras and small digging arms, and resident intelligent software enables the unit not only to identify weeds, but to conduct their removal.

How it works

Right now, we’re looking at sugar beet. The robotic units would be sent out into the field several times when the plants are young – that is, between sowing and the 12-leaf stage, when they are perhaps 20cm-30cm high, and plants are just beginning to touch each other. It’s at this point that the plants are at their most vulnerable, and when this technology is best applied, so the sugar beet can compete with the fast-growing weeds.

The units would straddle the seedlings as they travel down the rows, with the cameras logging everything they see and comparing it with an image database. Crop plants can be distinguished from weeds in just a tenth of a second, and differences can be identified even when there are overlaps.

These weed-removing-robots are a longer-term project for us. In the short-term, we’ll be introducing a weeding module, incorporating the intelligent software as well as the hardware, that European agricultural machinery manufacturers can integrate into their equipment.

It’s not just a proposition for organic farmers. Traditional producers also need to reduce their herbicide dependence: countries such as France and Germany are introducing restrictions over the next few years.

Looking to the future

In the future, our technology could perhaps be used to feed and irrigate sugar beet, too. It could also be applied to other crops, such as corn – but as with sugar beet, we’re using our robots to acquire in-field images of other crops to build a fund of knowledge for identification purposes.

It’s set to be an exciting year for us. When the add-on module has been launched, and the concept is proven, we’ll be able to further develop the autonomous vehicles. It’s great to think we won’t be just solving practical problems for farmers – we’ll also be helping organic and traditional agriculture alike to grow and harvest in a better and more environmental way.

Source: Rexroth