Precision farming: Leaders in the field

The day-to-day needs of a growing global population may vary, but everyone needs to eat and feeding the world is going to put a huge strain on the world’s agricultural resources.

The United Nations Department of Economic and Social Affairs predicts that the current world population of 7.6 billion will grow to 8.6 billion by 2030 and 9.8 billion by 2050.

According to Clive Blacker, director of Precision Decisions, a York-based supplier of precision farming services, there are estimates suggesting that crop yields will have to increase by 65-70 per cent in the 32 years to 2050. This could be problematic. “In the last 10 years we’ve hardly seen a five per cent increase in productivity,” he said.

One of the reasons for this is the size of agricultural machinery, which has steadily grown bigger. A significant proportion of farming is weather-dependent, which gives farmers in Britain unpredictable weather windows. Bigger machines get jobs done more quickly, and they also help compensate for the reduced numbers of available rural staff.

However, the trend to go big has had an adverse effect on soil through compaction, a situation that occurs when the weight of farm machinery literally squeezes the life out of soil by reducing its ability to hold water, nutrients and air.

Farming is going to have to get smarter if it wants to boost productivity, and doing this will require the deployment of a range of technologies to make farming more precise, sustainable and profitable. A research team led by Kit Franklin, an agricultural engineering lecturer at Harper Adams University in Shropshire, set out to do just that with the Hands Free Hectare (HFH) project.

Funded by Innovate UK and Precision Decisions, HFH is a world first proof-of-concept project. It started in 2016 with the aim of demonstrating that there are no technological barriers to fully autonomous field agriculture.

“[Franklin’s team’s] approach was for something that was open source and could challenge the current thinking,” said Blacker. “They wanted to use existing off-the-shelf components that would challenge some of the thinking of some of the bigger manufacturers, in terms of trying to make products more available without people having to buy into a system that they might not want on everything – which is typically the way some manufacturers are looking to go. They want ownership of everybody’s machines and data, which is becoming – at times – claustrophobic and restrictive.”

The initial one-year project was undertaken with smaller machines that included a lightweight, 38HP Iseki tractor for spraying, drilling and rolling a crop of spring barley. One of the project’s aims was to facilitate precision farming via automation with smaller, lighter machines that remove compaction issues and provide much greater resolution in relation to the application chemicals.

Jonathan Gill, mechatronics researcher and unmanned aerial vehicle pilot, flew regular drone sorties over the crop to acquire multi-spectral NDVI (normalised difference vegetation images) that would help inform agronomist Kieran Walsh, of crop production specialist Hutchinsons, about the condition of the crop and where to send a ground rover to collect plant and soil samples. The ground rover – a modified wheelchair – was also able to send video footage, giving Walsh more information about crop conditions.

According to Martin Abell, an engineer at Precision Decisions, this information allowed the team to apply fertiliser very precisely. “It’s all about putting the right product in the right place at the right time,” he said.

“Instead of putting a flat blanket rate across the field you vary that according to what the crop needs. It’s essentially being more efficient and more sustainable.”

When it was ready, the crop was harvested by a 25-year-old Sampo combine with a two-metre header unit. The team instructed and observed from ‘mission control’, a hut at one end of the hectare, but all tasks undertaken in the field were carried out by readily available machinery; open source technology; and an autopilot from a drone to help with navigation.

“This whole project revolved around taking the computer from a drone and placing it on the vehicles, so that the autopilot was then in control of each of the vehicles,” said Abell.

“We had to work out how to convert the signals that would [translate] into the movements that the human operator would normally do. That revolved around linear actuators and electronic motors, and they were mounted onto the conventional controls. Then we basically used motor drivers and different feedback mechanisms to work out what they were doing and interpret those signals from the drone autopilot into movements.”

With the drone autopilot installed for navigation, the tractor could then follow a predefined route in the field, making its way between waypoints, which are digital GPS markers positioned at the ends of the field for the tractor to navigate to. During the drilling phase, the waypoints incorporated lifting and lowering signals that picked the drill up at one end and placed it back down once it had turned around.

The tractor was first in line to undergo modifications for autonomous operation and, whilst the team was keen to prove open-source technology, it found also that systems designed for non-farming applications did not always translate in the field.  “When you put a GPS receiver on top of a tractor – with a few metres of height and a lean angle – then suddenly that makes an offset in your GPS,” said Franklin. “As the tractor would lean and wobble the programme within the autopilot software would chase that result, ending up with our tractor S-ing its way along the field.”

The spring barley yielded 4.5 tonnes, missing HFH’s predicted yield by just 0.5 tonnes. In November 2017 the HFH team received a boost from the Agriculture and Horticulture Development Board (AHDB), enabling it to embark on a second crop of winter wheat. The Harper Adams team has also been chosen to participate in RuralFirst, a UK government 5G project led by Cisco and backed with £4.3m in funding.

For HFH, mission control was receiving information from field vehicles via Wi-Fi, which had range of 150m. 5G promises nationwide coverage, and would give the HFH a range of benefits including improving the video link between mission control and the ground rover, which Franklin describes as “grainy”.

“At the moment we have a radio connection for our video and it’s all bit…crackly and we’re hoping that with 5G we can potentially go to four or five full HD video streams,” he said.

In February 2018 the Government announced £90m of Industrial Strategy funding to investigate the application of technologies such as AI, robotics and earth observation in agriculture. This, in turn, is expected to help boost rural economies and create new, high-skilled jobs. As Franklin noted, agricultural automation has the potential to boost jobs, rather than remove them.

Willingness to invest in precision farming is welcomed, but Franklin, Abell and Blacker all agree that the approach has to appeal to the very people it’s aimed at, namely farmers. “Farming is business and if you can offer a farmer a business case they’ll be interested,” said Franklin.

“Precision farming as a concept has been spoken about for the past 20 years and the reason for the slow uptake is because not much of the technology has had that really clear-cut business case attached to it.

“For things like spot spraying the maths becomes very easy. If I’ve put 80 per cent less chemical in my tank I’ve saved 80 per cent of what is one of my biggest current costs. That’s where it [precision farming] will become easier.”