A coastal Norfolk farm has become the test bed for a project aiming to get unmanned aerial drones and ground-based robots working together in the countryside.

Deepdale Farm, near Burnham Market, is hosting trials by agri-tech firm Antobot and robotics researchers from Loughborough University.

Their goal is to explore the combined potential of automated drones and ground robots in agriculture by developing a dock on the terrestrial machine so its aerial counterpart can land, transfer data and recharge - all without human intervention.

This could maximise the survey abilities from ground-level crop-sensing cameras and "eye in the sky" imagery, while extending the flight time of drones by recharging them in the field.

Zoe Stockton, business development manager at Chelmsford-based Antobot, said: "We have prototyped this docking mechanism, and we have been out to Deepdale to test it to make sure it works - which it does, so we are really excited about it.

"Eventually, we are hoping the robot can act as a portable base station, so the drone can land and be kept in place so it won't fall off and get damaged. Also it can charge the drone, which is important when you have a large machine with multi-spectral cameras, its flight time is limited.

"When it comes to combining the drone and the robot we are trying to make it more efficient, by combining the cameras on the robot and the drone to get information from two different planes. That can be useful if maybe the drone spots an issue in the field and the robots goes to take a closer look."

The company has developed a scouting robot named Insight, primarily developed to survey fruits such as strawberries or apples.

"It uses AI (artificial intelligence) algorithms to autonomously detect the fruit and process data on the size, count and ripeness," she said. "We have also started work on another robot platform which is more suited to arable crops.

"We are interested in exploring that angle and how we can adapt our platform so it could be used for different farms - perhaps to scout field vegetables or carry out robotic weeding, which could be particularly useful for organic farms that cannot use weedkillers."

In Deepdale's exposed coastal environment, a manual trial of the docking mechanism was a success - but there is still more work to be done to fully automate the process.

Dr Jingjing Jiang, a lecturer in autonomous vehicles at Loughborough University, is working on the auto-docking algorithm and control system for the drone.

"It is quite hard with strong winds in places next to the sea like Deepdale, but we want to challenge ourselves," she said.

"We did some initial testing for the manual docking. We had a drone pilot with 10 years' experience trying to dock the drone on the docking system, but he failed twice before he finally succeeded. So you can imagine how difficult it can be for a manual docking manoeuvre.

"But compared to a human pilot, we do have some benefits in using robot docking, because the drone has got a much clearer view with a camera facing directly downwards, but a pilot has to stand a distance away and their eyesight and viewing angle is limited.

"Hopefully by the end of this project we will be able to demonstrate auto-docking, but it is quite challenging."

Nathan Nelson, estate manager at Deepdale Farm, said it was exciting to offer "a real-world testing environment" to get these new technologies working in synergy.

"There is definitely interest for us as a farm," he said. "We could be looking at using an agricultural robot on the farm, whether it is weed management or the photography and imagery that we need to detect weed issues.

"If we can take diesel-guzzling vehicles off the field we can reduce the compaction of the soil as well, which is in keeping with our regenerative farming practices and trying to reduce our carbon emissions."

The project is supported by Defra and UKRI (UK Research and Innovation), whose funding also helps farms to host trials and accommodate researchers.