The Hands Free Hectare (HFH), a project run by Harper Adams University and Precision Decisions, a Map of Ag company, started in 2016 growing a crop of spring barley autonomously.
Three years on and two successful cropping cycles later, the project is set to expand from a humble one hectare, to the autonomous farming of a 35ha farm.
The initial project saw no-one enter the 1ha plot, with all work carried out autonomously and agronomists interacting remotely with data gathered from the field.
This in itself posed many challenges the team needed to overcome, challenges which have grown with the ambition and scale of the new project.
During 2019, the consortium has focused on the difficulty of moving vehicles around the farm.
Martin Abell, mechatronics engineer for Precision Decisions, says: “We have looked at technology from on-highway applications and evaluated whether this would be appropriate to move the tractor from the shed to the field.
“Many of the existing proprietary solutions do not work in this application. One issue is that they often require 5G connectivity and it is questionable which will be in place first – driverless technology or 5G?
“For other applications, the speed of connection is the main drawback, but in rural locations coverage is more relevant.
“If you are constantly having to monitor the tractor, you may as well be driving it.”
At present, the transport system is GPS-reliant, so automation begins outside the shed.
The consortium is also studying systems to complement yard to field transport and is trialling the use of ‘connected’ gates from manufacturer Dofygate which close off an area to pedestrians while the tractor is in it.
Using a GPS map, the tractor sends a signal to the gate when it has left the area so that it can be re-opened.
In addition to GPS guidance along a predetermined route to the field, the tractor itself is equipped with a number of safety systems to prevent collisions.
A laser scanner (LIDAR) is mounted on the front of the bonnet with two trigger points at four metres from an obstacle forward speed is reduced, and if the obstacle is in position as the tractor gets within 2m, it is brought to a stop.
Currently this is a single-plane laser, but will ultimately be replaced by a multi-plane stacked LIDAR systems to pick up lower obstructions, such as small livestock, operating in 3D.
A pair of ultrasonic sensors are also mounted on the sides of the tractor to detect objects approaching laterally, and more will be added, giving a more all-round view.
Bumper bars are fitted on the front of the tractor to push down vegetation and avoid unnecessary triggering of the LIDAR system.
It is accepted that these are not appropriate for some vehicles, such as harvesters, and other systems are being considered.
In its early incarnation, the tractor would stop rather than slow in the face of an obstruction and the team was reliant on a safe stop button to kill the tractor engine if it needed to stop to avoid an obstacle.
The operator-controlled button has been retained to give a final failsafe in case of emergencies.
ABOUT HANDS FREE FARM
FOLLOWING on from Hands Free Hectare (HFH), additional tests of autonomous technology will be provided on the Hands Free Farm (HFF) on the Harper Adams campus, a three-year project run in partnership between the University and Precision Decisions, along with a new partner, the UK division of Australian precision agriculture specialist Farmscan AG and funded by Innovate UK.
The Agricultural Engineering Precision Innovation Centre (Agri-Epi Centre) is providing the team with development space and project management support at their Midlands Agri-Tech Innovation Hub, which is also located on the University’s campus.
The 35-hectare HFF will have six fields linked by tracks and lined with hedges and trees, and autonomous vehicle communications will use ZigBee personal area network technology, said to give a better range than WiFi without the use of repeaters.
Up to three small tractors will be used for the project, including the original Iseki tractor modified for the HFH, and a Claas combine will be joining the original Sampo.
Martin Abell explains: “There will be plenty of opportunities to explore the potential of the technology over the three years of the project, which will start with winter wheat and move on to barley and other crops.
“We will look at how connected machines can work together as a swarm, for example, so if one tractor breaks down, another tractor working in the same field will know it has to plant every row rather than every other one.
“We have got lots of challenges to tackle, such as how to fill a small drill which can only cover a couple of hectares rather than going backwards and forwards with a loader.”
Kit Franklin, senior agricultural engineering lecturer, says:
“The HFH project produced more questions than answers, which is what we were expecting.
“We want the farm to become a testbed for agricultural innovation. Once the farm is established, we will encourage companies to come and test and evaluate their technologies.
“It is also great the project will remain on the university campus, so students will be able to learn from it, watch our progress and see how dynamic and innovative the agricultural engineering industry is.”
MR Abell comments HFH has already inspired student projects, such as the development of a yield monitor specific to the needs of the Sampo plot combine used to harvest the crop, and a penetrometer to work with the autonomous tractor and measure compaction.
“The ideal scenario is to be able to use the machine to its full capacity, but this could be prevented by failing to develop the technology or ensure any legislation applied to it is appropriate,” says Mr Abell.
Industry stakeholders have joined forces to initiate a code of practice for the agrobotics industry, farmer customers and the public to ensure their safe and sustainable future.
Representatives from technology and equipment manufacturers, the Health and Safety Executive, the Agricultural Engineers Association, Institute of Agricultural Engineers, funding body Innovate UK plus the HFF team has initiated discussions on indicators of a successful introduction of driverless machinery on-farm.
Clive Blacker, director of Precision Decisions, says: “If robots are to be part of the eco system on farms, there needs to be an awareness of how to ensure safety for operators and bystanders, and this must be focused on their use in agriculture.
“There are almost 700 standards applicable to connected autonomous vehicles [CAVs] according to the British Standards Institute, but we need to establish how agricultural machines fit into this legislation.
“We need to work out who is responsible should something go wrong.
“If we do nothing, there is the risk of the machines being classed in the same way as unmanned aerial vehicles which cannot be used out of the line of sight of the operator, which would be prohibitive to the development of agricultural CAVs.”
The group will now seek to develop the initial guidelines from the meeting and establish a representative body to engage with legislators.
Jonathan Gill says: “We need to be proactive and work towards standards which will allow this industry to progress for the benefit of farmers rather than having inappropriate legislation imposed on it.”