UK research trials of a new automated weed-killing robot have been able to reduce herbicide usage on crops by up to 95%.

A recent study at Reading University used cameras to identify weeds in vegetable fields. The technology – called eyeSpot – is programmed to target weeds individually and apply precise herbicide droplets with an ejector, which accurately fires treatment to individual leaves of each weed.

“This is precision agriculture in action, the robot has significantly reduced the use of herbicides, while practically eliminating any harm to non-target organisms,” said AHDB crop protection senior scientist Joe Martin.

Part-funded by AHDB, the research set out to protect the environment and help the industry manage with less access to crop protection products.

Nikolaos Koukiasas, a PhD researcher at the University of Reading, said: “EyeSpot represents a paradigm shift to weed control by accurately targeting leaf-specific droplet applications.

Preliminary results of manual droplet applications showed excellent weed control and a 95% reduction of herbicide use in cabbages, and 74% in leeks.

With big data becoming instrumental in farming, eyeSpot’s imagery also has the potential to be used for the observation of growth rates, enabling accurate scheduling of operations, early yield estimates and the detection of crop stress.

Alistair Murdoch, professor of Weed Science at Reading, said: “Yields and profitability are likely to equal or exceed those achieved by conventional herbicide treatments without applying any chemical to the crop.

The environment also benefits greatly by reducing the need for mechanical weed control, eliminating spray drift and reducing the possibility of chemicals entering the surrounding area.

Partners involved in the development of eyeSpot robot include Concurrent Solutions LLC in the USA.

Opportunities for commercialisation are being explored, and, if successful in later stages of trials, eyeSpot would be developed to serve the UK and potentially other worldwide markets.