The practice of farming is undergoing a change, one that many believe will ensure a sustainable future for the planet and those who reside upon it.
Janet Kanters speaks with Ben Scott-Robinson, co-founder and CEO, Small Robot Company
According to Ben Scott-Robinson, co-founder and CEO of Small Robot Company, cereal farmers around the world are in trouble. He posits that the system they use to product the world’s biggest crop is getting more expensive, while yields and revenues have flatlined.
“A recent UN report showed that we’ve lost nearly 200 billion tonnes of soil worldwide. Just when the world needs more food, we’re pushing our soils to the brink,” he notes. “This makes the next 10 years of farming existential in terms of what they [farmers] need to do. To fix this, farmers need a completely different model.”
During a virtual presentation at Farm to Fork Europe in June 2021, Scott-Robinson suggested that to do sustainable farming that can also deliver yields, a complete service needs to be designed from the ground up. And to make it work with millions of crop plants per field requires a radically different approach to that of the big tractor companies.
“Welcome to the fourth agricultural revolution – a radical change to the way farming is done, driven by a thousand start-ups that are bringing the digital first, AI driven revolution that is transforming our world to one of the last industrial era industries,” he says, adding that this change and these companies are asking farmers, food companies and governments to invest in a change as radical as when tractors replaced horses in the 1940s.
“The results will make the tractor as redundant as tractors made the horse then.”
Scott-Robinson says farmland is degrading: weather patterns are unstable, and yields have been wiped out by floods or droughts or fire. In addition, government subsidies and insurance are changing in the EU, the UK, the U.S. and Canada, to support carbon positive and sustainable farming.
“When you combine these additional revenues and counter these additional threats, over the next 10 years this revolution will stop the decline in yields and revenues, and allow farmers to make a profit knowing they are also making their farms sustainable for the next generation,” he notes.
Amongst these companies are hundreds of robotic start-ups. But Scott-Robinson says most of these are simply focused on delivering the automation of existing processes.
“While there are significant financial gains to be made by reducing labour on high value crops, it doesn’t fully take advantage of what the underlying technology can provide,” he notes. “And it doesn’t answer the existential problems facing the farming of the world’s staple crops.”
To that end, Scott-Robinson says Small Robot Company is delivering that next generation, a transformative breakthrough that they call “per plant” farming. This radical approach is focused on using lightweight, autonomous vehicles providing a near real-time view of each crop plant as it grows throughout the season, and then using this view to deliver precise and timely interventions to make that plant reach its full potential.
“By doing this, we can strip out all the inefficiencies and vast overuse of herbicides and pesticides, fungicides and fertilizer, and start using alternatives based around biostimulants or even removing liquids from the field,” he says. “And because we don’t disturb or compact the soil, we allow the ecosystem of the farm to do what it does best, which is fix itself.”
Scott-Robinson claims this saves 95 percent of the inputs, but also increases yields and locks carbon into the soil. “With this radical but deceptively simple vision, at a stroke, we can increase yields, increase nutrient levels within those crops, reduce the cost of production, regenerate soils and sequester carbon dioxide.”
Scott-Robinson says it’s not an individual process that needs automating, it’s a new model of farming, and with it a new way of providing what farmers need: a service that looks after the crop throughout its life and a suite of robots to deliver that. To do that, Small Robot Company has introduced robots Tom, Dick, Harry and Wilma.
“Tom lives on the farm continuously gathering data on the plants and the environment. Wilma is the brains of the operation, converting Tom’s data into instructions for Dick and Harry. Dick nurtures and protects the crop, he kills weeds individually with electricity and sprays only the plants that need their support. Harry precisely plants the crop at exactly the right depth and spacing for the conditions, giving it the best chance of the highest yields,” says Scott-Robinson.
Small Robot Company’s offering is somewhat different from other robotic and AI servers in that farmers don’t buy these robots – they pay per hectare for the service; they pay for the delivery of a healthy crop.
“We know all too well that to deliver one robot that works reliably in all conditions is tough, but three is a huge undertaking,” says Scott-Robinson. “Hence, we have developed a modular architecture across all our robots to allow us to be able to deliver them quickly and efficiently.
“We have already solved the hard problems first, the steering and power and control and suspension from the hardware side as well as autonomy and localization and awareness and electronics,” he adds. “By doing this we have derisked all future work onto the relatively simple task of chassis design and then bolting on components that we’ve previously developed.”
Thanks to the company’s completely unique view at a per plant level, they have developed their robots to deliver the realities of a field. Weeds, pests and disease are not uniformly spread, so they have developed a weeding system, for example, that’s not tied to following rows, but can delicately traverse the field to the place where they’re needed most.
“Our pride and joy is Tom. He is a third-generation robot designed to a farmer’s service requirements. He’s incredibly light, just over 200 kilos, and designed to cover a rough field while acting as a stable platform,” says Scott-Robinson. “We’ve partnered with the UK’s leading robotic manufacturer, Tharsus, to build Tom to a production spec. Tom’s capacity to gather per plant data is so unique that we’ve also approached five of the top 10 seed and chemical companies to help automate their research and development.”
Delivering this whole service at once would be impossible for a start-up, so the company is initially focused on showing the power of per plant farming through answering the most pressing existential threat for farmers right now, and that is controlling an increasingly resistant population of weeds with an ever-shrinking arsenal of chemicals.
“Our first service is very tightly focused. We will be delivering a system developed by RootWave (technology to create a fully autonomous weeding robot for arable crops) that controls these weeds using targeted electric weeding that literally blasts each individual plant with a mini bolt of lightning (electricity),” he says. “By using this system, we can offer rolling weed control at emergence, stopping any threats to the crop.”
Small Robot Company is providing the service to the audience that needs it most – farmers that are either organic or using zero till planting systems. Organic farmers need a way to control weeds that doesn’t need tilling which is obviously carbon releasing and damages the soil structure. And zero till farmers need a way to reduce their reliance on glyphosate and counter their overwhelming weed problem caused by leaving the soil disturbed.
“This is only part of the story,” notes Scott-Robinson. “By understanding each plant in the field, we’re able to bring a level of subtlety and efficaciousness to our system that is simply unheard of.
“On one side, we know that only 40 percent of the weeds in the field need to be completely eradicated; 20 percent can just be thinned; and 20 percent are actually beneficial to the crop plants. And of the bad weeds, we’re unique at being able to target the very worst – blackgrass.”
Blackgrass is so pernicious that it is taking entire fields out of production across northern Europe. Scott-Robinson says the weed is approaching 70 percent resistance to all herbicides and it is also visually indistinguishable from weeds until it is ready to seed, and the damage is done.
Blackgrass costs UK farming £400 million per year and 800,000t of lost harvest (ZSL, 2020). “It’s projected that the threat from blackgrass will grow into a billion-pound problem, just in the UK, and will be taking out a quarter of the entire wheat yield,” says Scott-Robinson.
Through Small Robot Company’s forensic view of the crop and using specific spectra, they are able to detect blackgrass as it emerges and then kill it.
“We’d love to say that this is our idea, but the reality is that we only came to this from listening to cereal farmers,” notes Scott-Robinson. “From the very start, we have been directly plugged into their fears and ambitions. The failures of trying to use the latest software with last century’s hardware, for example. The fears of a changing environment, of changing payment models and of changing consumer demands.
“We started Small Robot Company after six months of detailed qualitative research, and the very concept of farming as a service and per plant farming came directly from them. Which is why Small Robot Company so accurately knows their needs.”
Scott-Robinson claims their customers include some of the most influential arable farmers in the UK, the UK’s largest landowner, and a major supermarket chain. “To keep them involved, we’ve created three streams of engagement. Firstly, we have the 100 Club, a group of farmers who have signed EOIs to use our service when it’s ready. Next, we have created a training and networking program for forward-thinking farmers to show them how the business of farming will evolve in the world of the fourth agricultural revolution, and we call this Farm Ambition. And finally, we have a group of farmers who have prepaid for our service and who will act as our consultants; this is our farmer advisory group and they are our first customers.”
Small Robot Company’s mission is to help farmers feed the world while regenerating the planet. And their slogan? “Small. Is. Good.”
“You don’t have to do big to do farming. Accuracy and precision will allow us to be able to keep feeding the world for generations to come,” concludes Scott-Robinson. ●
Q: Do you feel that agricultural robots are compatible with this feeling of the farmer should be the final manager of the farm, or do you think that somehow this is taking the farmer out of the equation? A: What we’re trying to do with our farming robots is really provide a level of precision and detail which is just not possible for a famer at the moment. We work at a per plant level which means that we go out and monitor every single plant in the field, whether it’s a crop plant or a non-crop plant, and every single area metre-squared of soil. We understand the topography, and we like to keep an eye on the weather and the existing set-up for that. However, we’re not trying to take over the way the farmer runs their farm. What we’re trying to do is allow for data to be able to guide a lot of the decisions and then to be able to act on those decisions in a way that farmers can’t do at the moment. For example, with this per plant view, we can see when disease hits a field, when it hits the first few plants, and before they show/before they start spreading the spores throughout the rest of the crop plant, we can see that and tell a farmer and let them know this is happening and ultimately act on that very early on. So, we provide that level of precision which means that the amount of action that is required is much smaller. The amount of chemicals – if they’re using chemicals – is much smaller, and so gives a farmer much greater control. Ultimately, there’s an awful lot of decisions a farmer makes which are
very low-value decisions, and a lot of actions a farmer takes that are very low-value actions. What robots allow that farmer to do is to not have to do those anymore. By doing that, we’re allowing the farmers to spend more time to think about their strategic view of their farm, like what are they doing around the soil, what they’re doing around diversification, around vertical integration of production of crops or food or products, how they’re taking on board the big challenges around soil health. At the moment, farmers only have a small area of their mind space to be able to deal with that. We are planning to open that up so they can think about that much more.
Q. Robots, artificial intelligence, per plant data collection, that all sounds really expensive. Are growers and farmers actually seeing a return on investment when it comes to digital tech? And how long does it take from the deployment of digital tech on a farm until they start to see that return? A. Realistically, it’s a couple of years before a farmer will see a significant return on investment because we have to learn the field. Some of that is about going out and collecting data in the field over the course of a growing season, and a lot of it is also just working with the farmer to be able to draw out their knowledge and understanding of that field and that farm as we go along. However, the advantages are quite significant. If you think about the amount of cost that is taken up with chemicals on a farm for example, being able to use per plant intelligence to understand where those chemicals are deployed, or being able to keep a very close eye on what’s going on in-field on a day to day basis means that you can start removing those chemicals or moving much closer to a regenerative or organic model safely without there being a huge risk to crops and yields. Those return a significant return on investment – removing 90 to 95 percent of chemicals removes anything up to 25 percent of the costs on the farm on a year-by-year basis. Not having to buy/pay for tractors, drills, sprayers, etc., removes pretty much all of the capital costs on the farm. And that being replaced by a service which is paid for on a yearly basis and can be paid for in relation to the increase in yield is something that really fits the farmers pocket, rather than trying to force them to adopt new technology which they have to spend the money on and buy before they can see any value from it.
Q. You have Tom, Dick, Harry and Wilma. You’re focusing on a very specific question which is weeds that are really problematic for wheat. You also talk about a lot of different start-ups that are looking at robots in agriculture. Is the field going to get super crowded with all of these robots crawling all over it? Do you have plans to expand and make your robots more multi-functional so that they can solve other problems? What is the developmental pathway for this? A. Our ultimate vision is to provide an end-to-end service that does all the manual labour in the field apart from harvest. That includes the monitoring and providing the intelligence of what’s happening in-field. The planting, the precision application of whatever’s needed, whether that’s weed control, whether that’s chemical application, whether that’s nutrient application or stimulant application, and ultimately taking the crop right up to the point of harvest. For wheat and ultimately for arable crops in general, we don’t perceive there being a need for any other robots to be in-field at that time. Obviously, using good regenerative practices, you will be making your fields have a much wider and more varied rotation. So, if you want to move from that into greens or you’re looking to grow a veg at some point in the year, or you’re looking to grow some non-arable crops, then yes, other forms of robotics are certainly available. But over the course of the year, and for us, we predict that we will be the only people who need to be there. Our starting point is really focusing on the data intelligence we pull from the field. A couple of examples: at the moment it’s possible to get green area index, for example, which gives you a rough idea of how well your crop plant is doing in-field. We can give you the size, the number of leaves, and the variance on every single plant in the field, all, you know, 16 million of them if you’re talking about a 20-hectare field. That obviously allows you to start to get an idea about yields much earlier, you start to see the health of the plant and what should grow; you can also see variance, you can start to understand where parts of the field where it’s probably not worth putting any fertilizer down because you’re never going to get a decent yield out of that and you’re just throwing good money after bad; but also where you can add a little something, a little biostimulant, a little nutrient to be able to get more yield out of it. The other thing that we’re doing is being able to map and monitor all the weeds in the field, so blackgrass, for example, is our headline one because we’re the only people who can see blackgrass, and we can tell you were it is in the field within a month of it coming up; and as everybody knows, it is 70 percent resistant, and at the moment we offer the only way to kill it which is using electricity. We also see the other common non crop plants in the field as well, and what we offer to farmers is the chance to be able to only target those plants in the field that are actually going to threaten the crop. Not every plant that’s in a field that’s not a crop plant is a weed; that’s the whole thing about cover crops. But they naturally occur in your field as well. If you’ve got the chance to be able to leave the clovers that will grow there naturally or the meadow flowers that provide biodiversity and only take out the cleavers and the rye grass and the brome and the blackgrass, then you will have something that is closer to a meadow type environment which is obviously much healthier for the soil, actually much better for the crop plant and generally much better for the environment.
Q. Robots can coexist with the bees and the other pollinators… you talked about this key problem and market niche, about blackgrass which is a huge and costly problem for farmers and for which there isn’t another solution. Tell us about the research you conducted in order to find out what the needs of your consumers are? How long did that process take, and what does that say about the upfront investment that needs to be put into expanding digital technologies for other problems in the field? A. It’s a timely and costly process to work that out. Knowing that blackgrass is a problem is something that we’ve known from the outset. We are completely farmer focused, so we’ve been engaging with farmers since before we started to understand what that problem was. And for arable farmers in the UK and in northern Europe, blackgrass is a massive problem. We also knew that it’s a problem that is incredibly hard to solve. Blackgrass is visually exactly the same as wheat for the majority of its life, and it’s only really when it gets to the point of shedding seeds that you can really tell the difference between them. We were lucky enough to be associated with an organization in the UK called CHAP (Crop Health and Protection (https://chap-solutions.co.uk/) who do a lot of work with agriculture research universities across the UK. One of those was Rothamsted which is world famous for research that it does into arable farming generally, and they had done a fantastic piece of work by a PhD looking at how to use very specific spectra to be able to detect blackgrass plants early. The challenge with that is you need to look at the blackgrass plant as you go over it because it reflects different light at different angles. It has an electrified surface which means that as you sort of move your view of it, the light that reflects back changes, which is why it’s possible for farmers to see blackgrass when they’re walking through a field, but it’s very difficult to see from an image. So that research has allowed us to isolate those spectra to be able to create the identifiers – the AIs if you like – to be able to convert that into something that is visible. And because we have this really forensic sub-millimetre imagery of the crop plants, it allows us to be able to detect them very early. It’s not only a unique thing that we do, it’s only if you do what we do can you detect blackgrass in that stage. To do that, we’ve had to go through multiple rounds of testing, both in lab, in test plots, in fields, in fields in different lighting conditions, in fields in different parts of the country. Blackgrass is very genetically diverse, so we’ve had to look at different areas to try and get as much coverage as possible. All this stuff is incredibly long and involved and complicated to get right. The outcome is something that’s close to magic in terms of being able to solve a massive problem. ●
Photo: Small Robot Company
Small Robot Company’s “Tom” lives on the farm, continuously gathering data on the plants and the environment Photo: Small Robot Company
An innovative precision farming platform is employing artificial intelligence (AI) in the battle against weeds in grassland.
SoilEssentials, a precision farming solutions company based in Angus, Scotland, has harnessed expertise from the space industry along with agronomists and academics, to develop AI to overcome the green-on-green challenge.
SKAi (the SoilEssentials KORE Artificial Intelligence platform) has developed AI components which can be trained to recognize broadleaved weeds growing in grass crops and then take control of a crop sprayer to implement targeted control as the machine progresses through the crop.
“Green-on-green species recognition is probably the biggest challenge we must overcome if farmers are to benefit from the efficiencies of automated targeted weed control in the future,” said Dr. Gregor Welsh with SoilEssentials. “We are now able to train SKAi to target weeds in grass fields and automate the sprayer control via our machine mounted hardware.”
The new SKAi technology operates in real-time, meaning there is no need to pre-map the field; the on-board AI is set-up to constantly scan the vegetation, identify and selectively treat the weeds as the sprayer boom passes over.
“Trials of the system are showing high levels of success in the spraying of docks in grass fields and we are confident that our partnership will be able to progress SKAi until it can recognize and differentiate between a wide range of weed and cash crop species,” noted Welsh.
Graham Ralston, hardware director at SoilEssentials, said targeted weed control, resulting in reduced input costs and reduced environmental impact, is a win-win for farmers seeking more sustainable use of agrochemicals.
“There are many situations where blanket herbicide application across a whole field is undesirable – I’m thinking of broadleaf weed control in swards containing clover, for example. Ultimately, our challenge is to refine the technology until targeted control can be achieved even in what would appear to be difficult scenarios, e.g., blackgrass in wheat crops,” said Ralston. “And, of course we must also make sure that the system makes affordable economic sense versus existing spray application methods.”
SoilEssentials are partnered in the InnovateUK funded project by Deimos Space UK, University of the West of England and Scottish Agronomy Ltd. ●
Deere & Company has signed a definitive agreement to acquire Bear Flag Robotics for USD$250 million. Founded in 2017, the Silicon Valley-based startup develops autonomous driving technology compatible with existing machines.
The deal accelerates the development and delivery of automation and autonomy on the farm, and supports John Deere’s long-term strategy to create smarter machines with advanced technology to support individual customer needs.
“Deere views autonomy as an important step forward in enabling farmers to leverage their resources strategically to feed the world and create more sustainable and profitable operations,” said Jahmy Hindman, chief technology officer at John Deere. “Bear Flag’s team of talented agriculture professionals, engineers and technologists have a proven ability to deliver advanced technology solutions to market.”
Deere first started working with Bear Flag in 2019 as part of the company’s Startup Collaborator program, an initiative focused on enhancing work with startup companies whose technology could add value for Deere customers. Since then, Bear Flag has successfully deployed its
autonomous solution on a limited number of farms in the United States.
“One of the biggest challenges farmers face today is the availability of skilled labour to execute time-sensitive operations that impact farming outcomes. Autonomy offers a safe and productive alternative to address that challenge head on,” said Igino Cafiero, co-founder and CEO of Bear Flag Robotics. “Bear Flag’s mission to increase global food production and reduce the cost of growing food through machine automation is aligned with Deere’s and we’re excited to join the Deere team to bring autonomy to more farms.” ●
Precision Planting, LLC, a subsidiary of AGCO Corporation, has signed an agreement to acquire the business and assets of Headsight, Inc., a precision agriculture harvesting solution company.
Since 1998, Headsight has built a global reputation for designing and developing world-class, harvesting row guidance and height sensing technologies. Headsight’s flagship product is the family of header height sensors used in corn and grain harvesting. These products are known for increasing harvested yield, reducing machine damage and operator fatigue, and improving overall harvesting efficiency by accurately maintaining the height of the header above the terrain.
“Headsight’s success has been driven by their passion for improving farmers’ harvesting operations around the world,” said Justin Kauffman, general manager, Precision Planting. “We are excited about their continued global growth opportunities as they serve farmers of all brands of harvesting row equipment. Headsight offers an innovative set of capabilities that are a great complement to AGCO’s vision of being the trusted partner for industry-leading, smart farming solutions and enabling precision agriculture growth.”
Rich Gramm, founder and president of Headsight, noted the two companies’ cultures and goals are well aligned. “I am very pleased that Headsight is now part of the Precision Planting organization,” he said. “Their deep focus on anticipating and solving farmers’ challenges is the same approach that has made us successful over the last 20 plus years in building the Headsight business.”
Precision Planting intends to retain the Headsight team and maintain its operations in Bremen, Indiana (U.S.). ●
Israeli ag analytics firm CropX has acquired Netherlands-based Dacom Farm Intelligence, a crop optimization platform.
Adding Dacom’s crop protection capabilities, CropX now offers a comprehensive farm management platform with active irrigation technology, hardware-based soil data, fertilizer management and crop protection capabilities. The acquisition also gives CropX a new stronghold across Europe as it continues to actively expand its global footprint.
This is CropX’s third acquisition since the beginning of 2020. In January 2020, CropX acquired Nebraska-based CropMetrics, and in September New Zealand-based Regen.
“Dacom was the missing puzzle piece for our farm management platform – adding an established foothold in Europe, two decades of European crop data to our global machine learning engines, as well as a critical component – crop protection and recording – to our product mix,” noted CropX CEO Tomer Tzach. “We expect to exponentially expedite our global growth as we integrate our teams and technology.”
With the acquisition, CropX will establish its European office in Dacom’s current Netherlands headquarters and double the size of its global team. Dacom’s 3,000 customers, spanning both smaller growers and massive market players, add more than 20,000 farms in more than 40 countries to CropX’s portfolio. The deal also expand CropX’s reach into rainfed regions.●
AGCO Corporation, designer, manufacturer and distributer of agricultural machinery and precision ag technology, has launched the pilot of its Precision Ag Line (PAL) program, a tool designed to streamline support services for farming customers using AGCO solutions with mixed-fleet operations.
The company said that PAL makes precision farming expertise available to farmers using products from AGCO brands such as Challenger, Fendt, Gleaner, Massey Ferguson and Precision Planting, even when those products are used in
conjunction with equipment from other manufacturers.
Cody Light, field execution manager for AGCO’s Fuse division, said the program is directly aimed at ensuring its customers have the best farming outcomes and exceptional customer experiences with its products.”
“Precision farming techniques can become complicated, especially when products from multiple manufacturers are used to accomplish them,” said Light. “PAL puts farmers in direct touch with our technical experts who are very experienced at aligning disparate systems into effective processes.”
PAL leverages the AgriSync platform. The tool aligns with the dealer’s current support system – usually an existing customer service number – to streamline access to the service. Customers of participating AGCO dealers simply call or text a familiar number to speak with product and technical experts at the dealership or AGCO itself for immediate assistance.
PAL is currently being piloted with several members of AGCO’s North American dealership network. The company plans to expand the program to additional locations later this year and into 2022. ●
