The Hands Free Hectare project in Britain finds that many small robots may work better than larger, manned machinery
A study of autonomous farming at Harper Adams University in Shropshire, England, found that swarm farming with many small robots may be more efficient than farming with larger, manned machinery.
“The initial economic analysis suggests that robots could cut the cost of producing wheat in the United Kingdom by 10 to 15 percent compared to conventional equipment,” said Professor James Lowenberg-DeBoer of the university.
“Maybe even more interesting is the use of multiple small machines instead of large machines. This potentially allows for small and medium farms to be more cost effective.”
He said the scalability of swarm farming will be of particular interest in the U.K., where there are many small and irregular shaped fields that are hard to farm with large equipment.
The Hands Free Hectare project, which started in 2017 with one hectare, was intended to investigate whether the production of commercial crops with autonomous systems was feasible.
The project recently received funding from Innovate U.K. to expand its scope.
“This will be upgraded to 35 hectares with five different fields, which are irregular shapes and have some topography,” Lowenberg-DeBoer said.
A new challenge for the project is a traditional footpath running through a field that has to be considered.
There are footpaths in the U.K. that have been used for centuries, and court challenges have ensured that people have the right to walk on them, even on private land.
“It was decided it would be handled sort of in the same way that cattleman do when they have cattle in fields where there are footpaths. You put a sign up that says there are cattle in this field,” Lowenberg-DeBoer said.
“In this case, there will be a sign that says, ‘robots.’ ”
There are also utility posts, variable soil texture and irregular field shapes that the robots will have to navigate around, and they will have to share roads with other vehicles and people when they move between fields and the farmyard.
“These are autonomous vehicles that have a predetermined route through the field,” Lowenberg-DeBoer said.
“They’re autonomous but they’re not very smart. They have limited capacity to deal with obstacles.”
He said the project demonstrated it’s much easier to convert equipment that already has electronic controls compared to tractors that are purely mechanical, and that it was cheaper to robotize farm equipment than initially expected.
It cost the team approximately $17,000 to install autonomous technology into a 40-horsepower tractor.
“The overall investment in farm equipment may decline substantially with robotic equipment, in spite of the extra cost of making them autonomous, because you have the potential of using smaller equipment more intensively,” Lowenberg-DeBoer said.
Farm managers will also be less concerned about maximizing how much one human being can do because multiple machines can work in the field at the same time.
Winter wheat, spring barley and fababeans are direct seeded with a robotic tractor and drill, sprayed and then harvested with a small robotic combine.
“From an economic standpoint, and I’m the economist on the team, the fact that this is retrofitted equipment makes the calculations easier because we know what that equipment costs, we know something about the reliability and maintenance requirements,” Lowenberg-DeBoer said.
Test strips will also be farmed conventionally with larger tractors in the 300 horsepower range to see how the two production methods affect soil compaction.
“In the U.K., soil compaction is a really major problem,” he said.
“We don’t have the kinds of freezing and thawing that would be typical in Canada or in the northern U.S. Once you get it compacted, it tends to stay compacted.”
The team’s hypothesis is that lighter farm machinery will cause less soil compaction, which will have a long-term benefit of better soil health and enable higher yields.
He said the team used a 40-horsepower tractor because it fit the smallest conventional cropping equipment they could find.
“This is technology that is coming maybe sooner then they (farmers) expect,” Lowenberg-DeBoer said.
“If they think about how fast the adoption of auto-steer happened … there’s a chance that the adoption of either co-robotic or the completely autonomous systems could happen just as quickly.”
He said there will likely be multiple retrofit kits to make farm equipment autonomous within five years, such as SmartAg’s AutoCart, which allows combine drivers to operate a driverless grain-cart tractor.
Multiple specifically designed autonomous farm equipment will be on the market within 10 years, he added.
The research team’s economic analysis suggests that many fully autonomous systems will be deployed in the U.K. and northern Europe when autonomous technology becomes readily available.
“In the Canadian Prairies, U.S. Great Plains, Midwest, places like that, we’ll see more larger pieces of equipment that will be co-robotic, where they will be working with human beings in close association,” Lowenberg-DeBoer said.
This is because large rectangular fields are more conducive to larger machinery and there are less security concerns.
“With small equipment, like the little robotic vacuum cleaners and lawn mowers, nobody really cares. But if you think about a 300 h.p. tractor in Shropshire, if something went wrong and it went AWOL it would hit something,” Lowenberg-DeBoer said.