I’ve been reading recently about AlphaGo, the AI led computer which taught itself to beat human beings at the strategy game “Go”. Go was considered one of the most difficult games for a computer to learn; it is a game which relies on intuition and instinct as much as calculation. A computer became good enough at chess to beat a human being in 1997, when Deep Blue beat Garry Kasparov, but it took almost twenty years before a computer became good enough at Go to beat the best human players. It finally did it in October 2015 when AlphaGo beat Lee Sedol over a five game series.
There was one particular move that AlphaGo made, Move 37 in Game Two which took Lee Sedol by surprise - it was a move of such creativity and uniqueness that human observers thought it was a mistake. No human being would have made the move because it would have been considered outrageously risky. Only as the game developed did the beauty and brilliance of the move emerge.
As human beings, we thought we had achieved the peak of performance in Go, there was an established “best practice”. The AlphaGo computer suddenly showed the way to a new summit of possibility, one that had previously been obscured from our view.
At first glance, it may seem that a somewhat obscure board game has nothing to do with the future of farming, but I believe that it has everything to do with the future of farming. I believe that by changing the way that farmers work with computers, we will reach a higher level of productivity and creativity in the industry and that swarm robots are the key to unlocking this.
Before we look at the specific benefits of a farming system led by small, smart machines, let’s have a look at the what is the worst that could happen, if we don’t change the way that we farm. There are three main things that will happen, each of them disastrous in their own way.
The loss of healthy soils, leading to unviable farms
The loss of active ingredients, leading to a farming system which cannot use any chemicals or fertiliser
The loss of financially viable farm businesses, as increased machinery expenditure does not lead to increased profitability
The first thing that will happen is that, by continuing to follow a system led by big machines, we will see an ongoing degradation of our soil. Big machines are bad for our soil. In most instances, the use of big machines necessitates cultivation, which not only gradually kills off the organic matter in the soil, it also enables greater levels of soil erosion to the extent that our farms are physically disappearing. Around 9 million tonnes of soil disappears from the farms of Europe every year. There can only be one possible conclusion to this, which is that at some point our farms become unviable. Many experts think that this point may not be far into the future, certainly within the next generation or two, which is a terrifying thought.
Even farming systems which minimise or even remove cultivation still rely on heavy machinery, which cause compaction. Whilst there are some very good farmers operating some great businesses which minimise the impact of compaction, wouldn’t it be better to move towards an approach where we didn’t have any compaction, rather than always accepting it as a fixed component of managing the land?
The second thing that will happen if we don’t change anything about farming is the ongoing removal of active ingredients and fertiliser from our agriculture. In the system we currently use, it is impossible for us as farmers to be completely certain that everything we apply on our fields meets its intended target. By continuing to use machinery which has been designed to maximise speed of operation rather than maximise accuracy, we put ourselves in direct opposition with environmental groups, and ultimately large swathes of the wider public. Surely we have a shared goal - one in which we produce more food with as small an impact as possible on the environment. The machinery we currently employ is a key blocker preventing progress in this area. Organic farming in its current form is great, but if we could create a future in which we still used chemicals and fertiliser but much more accurately and far less wastefully, that would be a better result for humanity.
The third thing that will happen is that farms will eventually become financially unviable. Capital invested in machinery is failing to yield a return to farmers; machinery is getting more expensive each year, but it is not delivering a consequent return in productivity, quality or profitability.
There is a better way, and it is a way led by swarms of small, smart machines.
If we change our approach, and re-think our farming system, I believe we will see three key improvements.
An improvement in the physical health of our farms
An improvement in the level of data and understanding we have of our farms
An improvement in our philosophical approach to farming and the mindsets that we have as an industry
Replacing something big and heavy with something small and light has some inherent benefit. Less weight on the soil means less interference with everything that lives in the soil, which in turn leads to healthier soil and the protection of one of our most valuable assets. Farmers are acutely aware of this but constrained by the amount of progress that they are able to make by the hardware that is available.
We have developed a reliance on big machinery because of the emphasis on speed and scale. A system of swarm robots would allow the same level of speed of operation to be achieved, with lots of small machines collectively achieving the same coverage as one big machine, but with much less physical impact on the soil.
Soil is still an unknown frontier in many ways. Our understanding has of course improved over time, but soil scientists would admit that there is still a huge amount that we don’t know about the soil. We do know that when we damage the soil, we are losing an asset which holds enormous value to humanity in general and not just to the farmer. A swarm robot system will allow us to get a much more detailed view of what is happening in our soil, and how it is interacting with the growing plants. We will be able to create a digital view of the soil, in the first instance using measurement technologies that already exists but which are difficult, and boring, for human beings to use with the repeatable accuracy that is an inbuilt design feature of the robot.
As well as the soil, using swarm robots we will gain a much more detailed understanding of the growing crop, moving to a position where we are able to treat every plant in the field individually.
One of the key mental shifts that needs to occur when thinking about swarm robotics is in thinking of them as a direct replacement to the tractor. Farmers tend to think of robots as a smaller machine doing the same job as the tractor. This is the wrong way to think about them. Rather than thinking of replacing a single big tractor with lots of small ones, a better way of thinking about it is that we are unleashing a network of interconnected computers onto our land, collectively doing a job with so much more accuracy and with a such a greater level of understanding than anything we have had previously, that they are effectively doing a completely different job.
Potentially the biggest change that I see as a result of the implementation of a farming system led by small, smart machines is a change in the way that we think about farming and food production.
We will be able to farm in a way that is much more accurate and much more detailed and, potentially, completely different.
This is where the link back to AlphaGo comes in for me, because just as the AlphaGo computer was not constrained by the human notions of “best practice”, so will swarms of AI led smart machines on our farms potentially unlock new ways of thinking about crop planting, crop care, soil management and other areas of farming.
This interconnected network of computers will be a source of creativity for the industry, and creativity will become a more important skill for farmers to develop.
Large farms will be able to start drawing inspiration from small farms, the part of the industry where I feel that the majority of the creativity currently sits. Swarm robots will enable us to get away from rejecting ideas because “they won’t work at scale”, because each individual robot will not be working at, or concerned with, scale of operation. They will eventually be able to achieve a level of accuracy on any scale which is currently only possible on smaller farms.
Swarm robots will enable us to stop thinking of the field as a uniform production facility and allow us to embrace, rather than ignore, the huge variety we have on our farms.
Our mindset as a farming industry has become constrained by the machinery we use. Because of the machinery, we only think in terms of scale of operation and through growing our businesses by doing more of the same. We act like we are at or near peak productivity from our soils, and we therefore focus on incremental improvements.
I think that there is more to come from farmers and from farming. I think this new network of interconnected computers that we call swarm robots will enable greater creativity in the industry, and I think that this is going to take us to the next level of productivity without increasing the damage we cause to the environment.