Developing skills to suit a new farming system
by Arable Farming
Autonomous farming systems will require new skills just as much as they need new tools, suggest the researchers at Hands Free Farm, the innovative autonomous farming project being run by Harper Adams University, Shropshire. Jonathan Wheeler finds out more.
Hands Free Farm is the successor to the initial Hands Free Hectare project which successfully grew a crop on a one-hectare site on the Harper Adams University farm without operators on machines or agronomists in the field.
Hands Free Farm covers some 34.5ha and a number of fields and is aiming to grow a range of autumn and spring crops.
Designing a system that can operate such a unit autonomously requires new skills and systems, says Kit Franklin, senior lecturer in agricultural engineering and leader of the project.
It is also requiring fresh agronomic skills, adds Kieran Walsh, regional agronomist for Velcourt Advisory Services, who has managed the agronomy for both projects.
Last autumn, the Hands Free Farm team successfully drilled 22.26ha of winter wheat and 1.82ha of winter beans and at the time of writing (February 2021) the team expects to drill the rest of the site to spring oats soon.
During last autumns drilling season the tractor and drill worked autonomously from the moment they entered the field until they left.
That included setting into work and lifting out after every pass and drilling the headlands on fields with uneven shapes, including slopes and footpaths.
The eventual aim is for vehicles to work autonomously from the moment they leave the farmyard until the moment they return.
But their existing achievement marks a huge improvement in machinery control since 2016, when the original Hands Free Hectare project got under way, drilling on a level, one hectare site.
The teams long-term view is of a fleet of vehicles working and communicating with each other in the same field.
A first step towards realising that aim was taken at harvest 2018, when the combine offloaded grain on-the-move to a tractor and trailer running alongside it, with both machines working autonomously.
And the vision for this coming spring drilling season could include one tractor working the drill with another working alongside it rolling the crop.
For this autumn, autonomous working might also include the combine offloading grain to a fleet of tractors and trailers, with all machines working autonomously.
According to Mr Franklin, there is a clear division in the skills required, he says: The autonomous farm manager/ robot fleet manager of the future should only need the skills of a competent tractor driver of current times to control their machines.
Remember that inside a modern tractor cab is now a wall of computers with many programmable functions.
Autonomous systems will require set up and maintenance, as do current machines.
But one key aim of automation is to release the limited labour force we have on farms from driving so they can devote more time to higher value tasks, such as business planning and making agronomic assessments.
We have seen in the dairy sector that the adoption of robotic milking can provide managers with more data and give them time to analyse it.
This, in turn, has targeted their attention to where it is most needed, resulting in welfare and productivity benefits.
I believe automation can offer arable farmers similar benefits. Creating those autonomous vehicles and systems will require fresh skills, he adds.
Off-farm we need a whole new generation of agricultural engineers with far more electronics, robotics and communication skills. One of the original aims of the project was to highlight the variety of new, exciting opportunities available in farming and to attract students from a wider variety of backgrounds into the sector and that is now happening, says Mr Franklin.
The engineering courses at Harper Adams University have already adapted in the last decade to include a greater emphasis on electronics.
But we will always need the mechanical platform to be well engineered, so mechanical engineering remains core.
Among the new technologies being used in Hands Free Farm are drones and the increasingly sophisticated cameras they can carry, the drone activity being run by Jonathan Gill, Harper Adams mechatronics and unmanned aircraft systems researcher.
These technologies enable Mr Walsh to remotely monitor any weed and disease threats the crops at Harper Adams are facing, which saves significant travel time.
We have upgraded the equipment and monitoring systems used on Hands Free Hectare.
We are using a range of software to help monitor crops with NDVI , vegetation maps and weather information, he says.
One important advance has been in the speed at which he receives information.
In the past it would have taken several hours for images to be processed and get to me, now we can do that in a few minutes, he says.
The improvement in drone cameras has made weed monitoring much more efficient and effective, with Mr Walsh able to identify key GPS points so drones can be directed to take photographs of areas that need close examination.
But disease monitoring can be harder, which is why the project team is developing a full weather and soil monitoring system, with a main weather station already in place and soil moisture sensors being deployed.
Monitoring for disease has always been trickier than for weeds.
But we know weather conditions can drive disease infection, so accurate weather data is very important.
We will have a network of soil moisture probes which can send regular updates to me at home.
With so much wet weather in the past few years, these probes could also help assess what field operations might be possible.
Knowing the weather patterns accurately is important because we know the conditions that favour problems such as rusts and septoria.
Septoria on wheat is hard to detect in its early stages.
It is not visible and can remain in a latent stage for 28 days, so the true pressure isnt apparent until you can see the lesions on the leaves.
If we need a more detailed look at our disease levels we can use the new Bayer Rapid Disease Detection technology, says Mr Walsh.
The team is still working on ways of adding a similarly accurate pest detection function to the system.