The Hands Free Hectare (HFHa) team has hosted a discussion consider what a good practice guide for agricultural robotics should include.

The HFHa project was run in partnership between Harper Adams University, Precision Decisions, Farmscan AG UK and the Agri-EPI Centre.

Opening the day at the Agri-EPI Centre Midlands Agri-Tech Innovation Hub, Clive Blacker of Precision Decisions said: “The HFHa project was run in a very controlled environment.

“While working on the original project of growing and harvesting a crop, we realised that there would be the need for a system to navigate from the workshop or shed to the field, in possibly larger vehicles than we’d used in the project, and the journey would include travel on roads and a number of other obstacles.

“We secured some Innovate UK funding so we could develop our ISEKI tractor to be intelligent enough to drive itself to the field and to manage obstacles encountered en-route.

“We started by looking at what was already available and found that current systems for these tasks are in their infancy, so we’re not at a point when we’ll see them being fully integrated.

“We wanted to disseminate our learnings and look at questions today that will have to be answered if robots are to become an integrated part of systems; which I believe they will be.”

Blacker explained that he envisioned a framework would need to be created to include information relating to health and safety, industry standards and responsibility.

Senior agricultural engineering lecturer Kit Franklin said: “Previously our tractor had a fail-safe system which was very risk-averse; if it identified an obstacle it would just automatically stop and turn-off.

Now, it slows down if an obstacle is four meters ahead and pauses if something is one meter from its wheels. It’s a step-up from where we were, but we’d have liked to have gone further to avoid obstacles.

Setting standards

Blacker also highlighted the risk of tractors requiring two different systems.

“The last thing we want is to see two different robotic systems being required in a tractor due to the different safety standards for road and agriculture. We, therefore, believe that we need to create a good practice guide,” he said.

Kit also shared some of the findings from an economics report based on the HFHa written by Harper Adams University Prof. James Lowenberg-DeBoer, the Elizabeth Creak Chair of Agri-Tech Economics.

For the report, it’s assumed that human supervision of robotic labour is required just 10% of the time the machine is in the field.

“The study is the first rigorous economic analysis that supports the hypothesis that swarm robotics will dramatically alter the economic environment in which UK arable farms will operate.

The model shows that autonomous farming with small-scale equipment, as used in the HFHa, has the potential break-even at 100ha – which is equivalent to 500ha with contemporary UK systems in wheat production, increasing the viability of small farms.

Martin Abell, mechatronics engineer for Precision Decisions, said: “HGV drivers and other industries which require the use of large vehicles are heavily regulated. While this is not the case in agriculture; and it can be just as dangerous.

“Farmers are working extremely long, and sometimes unsafe hours, and we predict that with the help of autonomous vehicles the not uncommon 100+ hour-working-week should become a thing of the past.”

Discussions for the guide looked at three different areas associated with good practice: Society; the farming community (the end users); and the agricultural robotics industry.

The group believed that the perception towards robotics needs to be challenged; changing it from seeing it as a threat, to seeing the benefits they bring by carrying out tasks that no one else wants to do.

They imagine that a code of practice for robotic operations would not have legal standing but would rather provide an industry approach which could inform and support lawmakers in due course.

The debate around responsibility was highlighted as an issue that would need to be addressed; if a robot causes damage, who would be held responsible?