We’ve been interested in the field of agricultural robots for some time now (see what we did there?), so when we brainstormed who best to consult on the matter, there was only one person for the job: Simon Blackmore. Professor Simon Blackmore of Harper Adams University is the utmost authority on the design and build of agricultural and horticultural robots.
We had a really lengthy chat with Simon about all sorts of important issues – ranging from the big questions such as how to solve the food shortage crisis we’ll face mid-century, right down to the nitty gritty on farming and its future. The interview was so chunky that we’ve split it into two parts.
Tharsus: How do we feed 9.1 billion people by 2050? This question often attracts polarised views. Tackling a problem of such magnitude calls for the application of a range of methods, systems, and technologies – selecting the most appropriate for the situation. The adoption of existing knowledge can improve agricultural productivity plenty, but it is now widely accepted that R&D investment in new agricultural technologies like robotics is critical to satisfy the future demand for food sustainability. How much is the introduction of robotics and autonomous systems part of the solution to making cultivation sustainable? Do you feel there’s enough investment going on at ground level to support this?
Simon: There’s investment going into things like plant breeding, biotechnology and chemicals. The thing that hasn’t changed much in many years is the investment into new mechanisation systems and so a lot of the solutions that we have now are heavily reliant on chemicals.
British society doesn’t like chemicals being used on our crops and foods because we think that chemicals are a bad thing. There is an increasing drive now by the environmental lobby to reduce and/or to ban more and more active chemicals. The interesting thing is that everybody is concentrating on these active agents, but in fact, there is nothing wrong with them. The problem is with the machine that applies them. The problems that occur are when all of these chemicals are applied off-target – so when farmers are spraying the whole field and only 10% of the spray actually goes onto the crop, or 1% when you’re talking about pesticides – then we’re getting huge wastage and it’s all of that wasted chemical that’s going into the ground water and causing the problem, but it’s not the chemical itself it’s the machine that’s applying the chemical that is the problem.
As well as monetary investment, there’s also a need to invest some time into educating different communities surrounding the agricultural industry:
- The farmers who are at the heart of this sector need to see with their own eyes how easy, adaptive and efficient these new high-technology products can be. How strong the economics are that support them. The only way we do that is through demonstration.
- This industry is starting to get a lot of press now – the rate of change is extraordinary, so getting the message out through the press is really important.
- The Government (who are key beneficiaries of the agricultural industry) are starting to open up the lines of communication. They’ve recently published a report built on the evidence I gave at the House of Lords a couple of months back about autonomous vehicles, so bit by bit we’re trying to raise the government’s awareness of the opportunities here. In fact, one of the outcomes highlighted in the report was that “robotic agriculture should be supported more by the government so long as the business case is made”, so we are now preparing that business case to approach the government to make sure that this area is fully supported.
Tharsus: According to the International Federation of Robotics, agricultural robots will be the fastest growing market segment of the robotics industry by 2020 – with 90% of that market residing in the EMEA. From an industrial perspective how does the UK fare in the development of these high-technology products in comparison to the rest of the EMEA, Americas and Asia Pacific?
Simon: It’s up to us really. This morning I was updating a list of robotic companies making agricultural robots. There are 10 companies of specific interest on my radar that are making agricultural robots now. The interesting thing is that half of them are still only half-baked. Their approach is a little off; thinking that they just need a small machine and that everything else needs to stay the same. It’s not fully thought-out. Most of these are built by people who are interested in the technology and not the solution, and it’s this mentality that will ultimately lead them to failure. As engineers, we like to think of how we’re going to solve these problems, and that’s great! Being a great engineer is at the core of what we teach our students here, but it really does require a better understanding of the environment, it’s a lot more complicated than just coming up with a new machine and then trying to sell it. You’ve got to work out how it’s going to be used, how it’s going to be managed and get all of these machines to work together as a system.
I could be thinking too far ahead in the future but I certainly consider all of my work and the work we do here at Harper Adams to fit into a complete system rather than a single solution. I can give you an opinion with regards to the UK’s position in adopting high-technology products in agriculture as I’ve been involved in precision farming for many years now – 20-25 years to be exact. I’ve spent a lot of time looking at the adoption of technology in this sector, specifically: why have some been successful? And why have some failed?
First off, the whole of Western Europe has great technological support. We have skilled people and the ability to deal with high-technology. Secondly, there’s a lot of talk at the moment regarding the implications of Brexit; the reduction of seasonal labour and the devaluation of the pound. There’s a range of reasons why the farmers who use seasonal labour are very nervous now. I know of some farmers who are reducing their cropping area this year because they don’t think they’re going to get enough people to come and harvest their crops next season. One thing British farmers all have in common is that they are experiencing serious economic pain – this is inevitably going to slow down industry innovation and the adoption of new technology in the short term. I was speaking to a grower a couple of weeks back who said that he spends £200,000 a year on paying people to do mechanical weeding on his farm. He’s paying people to go into the fields and take out weeds by hand! To me, there’s an opportunity straight away. He has said that if he could get a machine that costs £100k he would buy one today. The economics are just there! The UK is probably one of the most advanced countries across the world when it comes to the acceptance and adoption of high-technology products. Nervousness about our future in the EU is going to dampen progress over the short term, but we have access to all of the right talent and skills that will allow us to move beyond that.
Tharsus: Total UK income from farming in 2015 was 29% lower than 2014, a fall of £1.5Bn to £3.7Bn in real terms. The picture across the developed world is similar. With the cost of adoption in robotics still relatively high, farmers need to believe in the long-term benefits of the technology and maybe even take a risk in deploying it. For what has been known as quite a conservative industry and given recent disappointing incomes, do you think farmers are in the right frame of mind to invest in robotic technology?
Simon: Part of my work looking at why farmers adopt high technology in the UK is to do with economics. If a farmer has an expensive problem – mechanical weeding or black grass (which is a real problem throughout the UK – because they are now becoming herbicide resistant) – I’ve heard of them ploughing up complete crops just to get rid of these weeds, which is crazy! Farmers will adopt high technology, but it requires three areas to be satisfied:
First off, it’s got to have embedded economics, so let me give you a good example of a piece of high-tech equipment that is going to be very successful because of its commercially robust economics. “Autosteer” on tractors is an investment of between £15-20K and gives farmers a positioning accuracy of 2cm. This provides farmers with precise data on where their tractors have been and minimises the skip and overlap of applied treatments. Usual savings of employing this technology are between 10-15% on chemical expenditure, so many farmers can gain pay-back for it within just a few years, so they’re quite happy to invest.
Secondly, new technology needs to make the farmer’s life simpler. So in the early days of my precision farming work – the technology we created made farming more complex. So on that basis, many farmers rejected it. Agricultural technology has to be developed like any other piece of high-tech equipment, like your smartphone. It has to be developed to a level of sophistication on the inside that makes it revolutionary but it has to remain simplistic on the surface so people can just pick it up and use it without having to have a PhD and an engineering degree to be able to understand it. The psychology of it is very important.
Finally, it has to have the embedded knowledge within it to allow it to do the jobs that you want it to do – in a smart way. GPS uses satellites to pinpoint location, but it would be incredibly difficult to explain the several data and timing signals of this technology because of its embedded nature within our smartphones and our cars. But it’s there and it’s doing its job. So if these three things are embedded within the product it will be successful. But it requires quite a level of maturity to be able to get it to that stage. First and foremost, it’s got to make the farmer’s life simpler, and do what it says on the tin, and then they will start to invest in it.
Tharsus: Bearing this in mind, do you think Robotics-as-a-Service (RaaS) could help lower the barriers to entry for technology providers and speed up the widespread adoption of robotics in agriculture? Are there any examples of businesses that you’re aware of that are currently exercising this model?
Simon: Definitely. This is one of the hot topics of discussion in industry at the moment. It’s not a product – it’s a service. There are great reasons for doing this, with a product you have to spend years conducting research, building the machines, testing them and so on. If you create products as a service, then the knowledge and the understanding of the product’s development can be done by the team in the field – in the commercial environment. Rather than farmers buying a tractor and having to understand the risks of the system and develop the expertise needed to run that tractor or chemical – they can just buy the service off-the-shelf.
So this model would allow a farmer to make an easy transaction (buying the absence of weeds for example) and give the service company the opportunity to be able to do a lot of in-field testing and own the captured data. The service company would then be able to gain an abundance of learning and experience; which is the right chemical to use, what’s the right dose rate, is this laser working, is this the right frequency, can we move faster, can we move slower. All that insight then resides with the people who need that expert knowledge – the service providers. It’s a classic model. If you look at precision farming companies in the UK there’s: Precision Decisions, Soil Essentials and two or three others. They all offer services enabled by high-technology products. The model is there and it’s accepted. I would highly recommend robotic technology creators to think about deploying robots through this kind of service model – because it gets products into market faster.
Tharsus: Exciting new field tests like the ‘Hands-free Hectare’ and the REAH project are starting to build the business cases for robotics in precision agriculture, primarily using off-the-shelf-technology. Will it be through these kinds of field tests that we really start to see a positive attitude change towards unmanned agricultural robots or will commercial organisations bring robots to market in parallel?
Simon: Yes, these field tests are really important because when you look at farmer behaviour – the main driving force behind why farmers are making buying decisions is not economics but is in fact an obsession with what his neighbour is doing – it’s like “keeping up with the Joneses”. He looks over the farm hedgerow and he sees a nice shiny new tractor or a special crop, or he sees something else going on next door. The farming community is tight-knit, so this is the fastest way to get farmers to invest in technology – if they see one of their peers investing in it and making money from it – then they’ll do it themselves.
What we have to do is get more and more of these demonstrations being carried out in fields. Firstly, to show what this new technology can do, and secondly to be able to reduce the risk in the farmer’s mind so that he can see the equipment working and say “well look it’s not so risky, they’re doing a good job with that, therefore I’m going to invest in it myself.
Tharsus: When proving the economics of widespread robot development in the agricultural domain, should we aim for complex robotics which aim to remove humans altogether (and so is a bigger prize in cost saving) but is harder to achieve and may cost more, or should we aim for simple robotics which make humans more productive and could be deployed more quickly and cheaply?
Simon: They’re two completely different classes really, one way more extreme than the other and arguably unachievable given the nuances that come with robotics still being ‘first of kind’ in nature. It’s unrealistic to assume a robot could be left to its own devices at this point in time. In a dystopian future, maybe…
First off, removing people completely. You could never do this, robots have to work in a supportive environment. I did an economic survey a few years back and found that the best opportunity for deploying robots was grass cutting, so it wasn’t agriculture at all – it was cutting grass on golf courses. If you’ve got a team of 8 or 9 people looking after the golf course, replacing the first person is very easy, but replacing the last person is impossible. So, you cannot have a robotic system without people as part of that infrastructure. What you’re doing is taking the person off the steering wheel and giving them other tasks to do. You’re moving the tractor driver into the robot operator.
The latter subset of robotic products is what is commonly referred to as the “co-bot”; robots working in unison with people to boost productivity. The MRAAV strawberry harvester is a great example of a co-bot whereby the robot is supervised by the person. The robot is then not only doing the selective harvesting but is also grading the strawberries so that we can replace the whole grading line in the factory. For the whole of one strawberry – you can use infrared and machine vision to grade it as a Waitrose or Tesco variant – and make a whole range of additional grading decisions like: shelf life, diseases, contamination etc. What they do at the moment is: humans pick the strawberries and put them into a basket, they then send them off into the factory through a production line where they are then manually making all of those grading decisions. You can replace a lot of these semi-skilled, highly repetitive tasks with the adoption of collaborative robots. So in summary, you can either have the robot working semi-autonomously – by itself but still with a person embedded within the system, or you can have a co-bot where the robot and the person work side by side, but you’ll never be able to remove humans altogether.