During the inaugural Future Agricultural Technology event, held at Birmingham's NEC late last year, one key message was clear: agriculture in its current state is unviable. Practices that have barely changed in decades typically result in stagnant incomes around the world. Many feel the industry is overly reliant on chemicals and machines that are zapping life from once-rich soils. But new technologies are helping the sector to reverse this level.
“We are part of a global trend in the way agriculture is changing,” said Sam Watson-Jones, co-founder of UK agritech startup Small Robot Enterprise (SRC). “We think arable farming in its contemporary form doesn’t work.” As a fourth-generation Shropshire farmer, Watson-Jones spoke from experience. He said yields have remained static for more than a quarter of an era, despite the UK using more than a million tonnes of herbicides and fungicides every year.
“The third agricultural reconstruction is what we are experiencing today,” he added. “Chemicals define it – it’s surrounded by fertilizers; is defined by large, heavy tractors. But it's not time-consuming work for farmers and we need something new to take us into the future. “The fourth agricultural revolution, on the contrary, may be determined by swarms of small, intelligent machines and minimal or no cultivation.”
Large tractors, sprayers and harvesters deployed on farms around the world are compacting soils and depleting nutrients. This is a challenge. Downsizing machinery not only preserves soil, but also allows workers to farm on a much more granular scale. This means that individual plants and crops can be tended to for better care. The process is called precision agriculture. However, smaller machines and “precision farming” typically mean more time involved in the process.
“We are moving away from considering our fields to collecting data and taking action on a specific plant – each plant in its field,” said Watson-Jones.
It's a method championed more than a decade ago by Professor Simon Blackmore of Harper Adams University in Shropshire, and which led Watson-Jones and co-founder Ben Scott-Robinson to launch the Small Robot Company (SRC) in 2017. Their robots Tom , Dick and Harry individually take care of field monitoring, weeding, cultivation and planting.
Of the three, “Tom” is in the most advanced stage of development and is expected to enter commercial service this year. Its latest iteration has been launched at Future Farming Technology, claiming a dual camera system that increases monitoring capacity to four square meters. According to the SRC, Tom can cover approximately 20 hectares per day, feeding individual plant data to an artificial intelligence system – identified as Wilma, which helps farmers evaluate the information.
“Robots are automating tasks, but “Wilma” is the brains behind the operation,” said Watson-Jones. “She is the part that will cause the change to a new form of agriculture.”
Together with Tom, SRC also unveiled a commercial service for weed mapping, a world first according to the company. Broadleaf weed heat maps allow farmers to use sprinklers more effectively and better shape future planting decisions. Eventually, “Dick” will also serve and remove weeds autonomously, microspraying based on data and analysis from Tom and Wilma. Harry, the agricultural robot, is still in the prototype phase. The three will work under a service package where the company will charge per hectare to distribute the land. This means that farmers will not be forced to make major initial investments.
SRC also operates a Warwickshire-based company called Rootwave, which uses an electrically energized metal arm to boil weeds from the inside, killing them from the roots up. It's all part of a master plan to transition from the mass use of chemicals to a much more targeted use – and, perhaps, one day, using zero chemicals.
Robot Tom can cover around 20 hectares of farmland per day, feeding individual plant data to an artificial intelligence system.
precision agriculture
“Reducing inputs is important,” said Jamie Butler, a dairy and arable crop farmer who has participated in SRC experiments. “Obviously for the agricultural economy, but it's also definitely the way forward for the environment, for soil health and management. I would say 90 percent of the chemicals and fertilizers we are incorporating into (our fields) are probably not needed and could be significantly reduced with the right technology.”
Butler agreed with Watson-Jones's sentiments about agriculture's broader problems. Despite remaining an early adopter of technology, he has had to diversify the Hampshire farm he operates with his brother. The duo offers glamping, self-storage homes and corporate fly fishing to supplement their regular income from farming and ranching.
“In some respects, I think the yields on farms are – maybe not as high as they can be, they could last longer – but I’m not convinced that farmers would see a profit in that,” he said. “We would just be supplying an oversupplied market. Why do we need to produce more? What we need to do is produce more efficiently.”
UK markets may currently be oversupplied, with large retailers squeezing farmers' margins to the limit. But the preservation of tomorrow's food is more fragile. It is estimated that the world population will reach 10 billion in 2050, around 2.5 billion more than today. If chemical-intensive agriculture and monstrous machinery continue, soils could eventually be pushed beyond the point of no return. Consolidated with the additional threat that climate change poses to land management, the status quo must change.
“I think precision agriculture will be a standalone revolution, with autonomous vehicles and also with weed testimony,” Butler said. “Right now, we’re not beyond that, but if people like Small Robot Company realize their vision, we’ll be there.”
Autonomous farming
The SRC is by no means the only agency in the UK researching autonomous robots as a solution to agricultural issues. Harper Adams University has been conducting a research project on autonomous agriculture, using small, retrofitted engines and harvesters, along with drones and AI.
Known as Hectare Mãos-Frees, the project harvested its first spring cereals in September 2017, followed by six tonnes of winter wheat a year later. The project continued for another three years, growing from an ideal hectare to 35 hectares in five different areas. The project has since been renamed Hands-Free Farm .
“These are fields that have not undergone any changes from standard agriculture,” Jonathan Gill, a mechatronics engineer at Harper Adams and one of the project leaders, told a popular UK publication.
“We have five fields, all with non-straight headlands. The most challenging one has telegraph poles, there is a public walkway in the center, it is undulating, there are four different types of soil.”
While SRC has specifically built robots from the beginning, the team at Harper Adams has taken a different approach, customizing current, compact machines with technology that allows them to farm autonomously. For Hands-Free Farm, a second Iseki tractor was added to the fleet, along with a Claas combine with a much smaller footprint than the previous combine. The team's partner, Precision Decisions, is managing route planning and control systems, with FarmScanAG working on autonomous capabilities.
“What we’re looking at is the level of smart implements, the implements and the machinery that goes into the vehicle,” Gill said. “The new cropping systems technology, the new bail technology and everything.”
For the original hectare, the team discovered the technology to plant, care for and harvest without any human behavior in the field. The Mãos-Livres Farm will have a more practical passage, agronomists and students working together with the autonomous machinery, collecting soil samples manually and making some old-fashioned judgments.
“I would never really want to have a farmer or agronomist going out on the land and making decisions, so let’s not do that,” Gill said. “What we would like to do is provide additional tools to help them do that.
“It has become much more sensible, there is much more professionalism behind the entire project to try to operate something routinely, instead of just being a feasibility study. It’s no longer about viability, it’s more about proving the capacity of autonomous agriculture.”
Part of this test will involve developing a realistic view of the economy and determining where autonomy can potentially generate savings. Data from all vehicles regarding race times, distance and fuel consumption will be collected and processed. For harvest data, new partner Pix4dFields has teamed up with us to provide a drone system that will capture regular updates above the fields.
Gill, a drone expert himself, believes UAVs will play a crucial role in the coming agricultural revolution, and not just for monitoring. He cites the example of XAG, a Chinese manufacturer of precision spraying drones that were deployed across two million hectares in China this year. It's a technique that is currently not permitted in the UK by the Chemicals Regulation Division (CRD), something Gill believes should change.
“A drone can fly at a lower height than a boom can operate and deliver a chemical with greater precision than most standard sprayers,” he explained. “A drone sprayer does not have the same precision as a single-nozzle controlled sprayer, but it is the middle ground, and the price of the technology is much cheaper and does not cause compaction.”
Drone spraying also unlocks the possibility of growing crops when the soil is saturated and farmers are unable to place a conventional sprayer in the field.
“There are perfect capabilities for these drones to operate and work in these environments,” said Gill, “but we are being hampered by our regulatory system that prevents us from actually operating these vehicles.”
Drones are ideal for controlling large crops like wheat and corn, but tracking different fruits and vegetables requires technology in the trenches. Mamut, produced by engineers at Cambridge Consultants, is a compact four-wheeled robot that travels across fields collecting and analyzing data. Using stereo cameras, LIDAR, an inertial measurement unit, a gyrocompass, wheel odometers and integrated AI, it can navigate new environments autonomously, offering a real-time picture of crop health at ground level.
“Fruit and vegetables – especially citrus fruit – you can’t see what’s going on before because the canopy covers them,” explained Niall Mottram, head of Agritech at Cambridge Consultants. “And there aren’t enough hours in the day for people to walk up and down the rows of an orchard or vineyard to count grapes or apples.”
Central to Mamut's effectiveness is its ability to operate independently of GPS or radio support, as well as its machine vision and AI that analyzes harvest data without the need for external computing.
“This kind of full AI, where you don't need to use a lot of power from the cloud computing platform, as that's not practical in an agricultural environment, you don't have real-time connectivity. This kind of approach is critical if you are working to see AI move out of the data center and into the field to generate some interest,” said Mottram.