← Business Daily
Business Daily
Is AI about to transform food production?
00:18:29
Farmers are increasingly using artificial intelligence-powered machines to boost business
Loading summary
Transcript43 lines
- [00:00]
A
This BBC podcast is supported by ads outside the UK. The best B2B marketing gets wasted on the wrong people. So when you want to reach the right professionals, use LinkedIn ads. LinkedIn has grown to a network of over 1 billion professionals, including 130 million decision makers. And that's where it stands apart from other ad buyers. You can target your buyers by job title, industry, company role, seniority skills, company revenue so you can stop wasting budget on the wrong audience. It's why LinkedIn Ads generates the highest B2B return on ad spend of major ad networks. Spend $250 on your first campaign on LinkedIn Ads and get $250 credit for the next one. Just go to LinkedIn.com Broadcast that's LinkedIn.com Broadcast. Terms and conditions apply. This message comes from Schwab at Schwab. How you invest is your choice, not theirs. That's why when it comes to managing your wealth, Schwab gives you more choices. You can invest and trade on your own. Plus get advice and more comprehensive wealth solutions to help meet your unique needs. With award winning service, low costs and transparent advice, you can manage your wealth your way at Schwab. Visit schwab.com to learn more.
- [01:17]
B
Hello and welcome to Business Daily from the BBC World Service. I'm Rob Young. Today we'll be glimpsing into the future of farming and how artificial intelligence could transform the way we grow food.
- [01:31]
C
This is going to completely revolutionize agriculture. Precision farming can essentially transform the bottom line of farmers across the globe.
- [01:39]
B
The promise? More food, using less water and with lower costs. But there are warnings too, as in.
- [01:47]
D
Any sort of data driven endeavor you've heard of garbage in, garbage out. And this holds even more true for.
- [01:53]
B
AI that is the cutting edge of agriculture. On Business Daily from the BBC.
- [02:08]
E
A.
- [02:08]
B
Farm in California that hasn't changed much in decades. Flat land stretching for miles. Rows of leafy green vegetables almost as far as as the eye can see. But where workers once bent under the hot California sun, putting up weeds and applying fertilizer, there is now a one million dollar machine powered by artificial intelligence. It looks a bit like a futuristic combine harvester. Third generation farmer Daniel Alameda says finding workers has been tough and this machine is doing a job nobody enjoyed anyway.
- [02:49]
F
So mainly what we've been using AI for is differentiating between a plant and a weed. And I know that sounds very simple, but it's actually very difficult to do. For lack of a better term, a weed is an invasive plant, something in that field that we do not want there. Why is that? Weeds compete for the same thing that a ahead of lettuce does. They compete for nutrients, water, sunlight, and they also harbor invasive pests. So we have to remove them from the field. So our biggest challenge is removing unwanted plants from the field that have traditionally been done with people. So we have to distinguish the difference between a plant that we want there and a weed that we do not want.
- [03:39]
B
And I suppose in a field of spinach, anything that's green looks like spinach.
- [03:44]
F
Correct. If all things are green in that field, it's hard to differentiate between the two. So when we started, more or less, we knew where we placed our seed. So we kind of knew where that healthy wanted plant should be. So we used a basic set of parameters to make a decision which was, I'd say 85% accurate for what we wanted to accomplish. Now to get that last 15%, you've got to distinguish with AI the difference, and that is plant and weed identification. So that's getting us that next 15%, which is really where the value's at. So that's where we've come over the last 10 years.
- [04:30]
B
Here's how it works. Cameras scan the field detect weeds, and robotic arms deliver micro doses of weed killer or fertilizer, supposedly with millimeter precision.
- [04:43]
F
Right now we've kind of been working with directional sprayers, which is applying either fertilizer or some conventional or organic chemical or laser. Each one of these AI driven weed removal machines kind of has a. It's not a one size fits all in each category. We kind of move them around to whatever's going to be best.
- [05:07]
B
And what is the cost effectiveness? Are you saving money using the AI automated weed killing system compared to the cost of getting people to do it manually?
- [05:16]
F
So that's what we're evaluating right now. And we don't think of these as labor replacements. We think of these as labor assists. Our workforce isn't growing, so we are trying to fill a void that currently exists. A lot of this tech is new. Some of these machines run from a half million to 1.5 million. As we expense that out traditionally over five years does that pencil. We honestly haven't been using these machines for five years yet, so we don't know, but it's looking like it's going to get close.
- [05:48]
B
And when you first saw the effectiveness of AI on your farm, what were your thoughts? Because farming is traditionally a manual intensive job and there is this almost romantic relationship between man and soil, they're not.
- [06:03]
F
So much revolutionizing what we're doing, they're kind of mimicking what we've already been doing and taking a repetitive task and assigning it to a machine. It's just the next step, right? It's allowed us to really focus on bigger picture items where we can just assign these machines this repetitive, redundant task of weeding and thinning. And it's made things a lot better. There's nothing romantic about a million dollar machine, right? But if it's doing the job effectively, like that's great, that fits in exactly with what we want to do. Costs are going up. We have to find these little niches that we can carve out to try to save every penny that we can. So, like these type of things are helping. Like, this has been the first time in my career in farming, which has probably been about 15 years, that that have actually seen something. You're like, this can cut our cost. Like, this will be effective.
- [07:00]
B
Daniel Alameda in California that is just one example from the cutting edge of farming tech. Other farms are using satellite data, drones and autonomous robots. Bhaskar Ganapati Subramanian runs the AI Institute for Resilient Agriculture in the US Farming state of Iowa. It's funded by the American government.
- [07:22]
C
My opinion is that this is going to transform and completely revolutionize agriculture. So what are the precipitating factors that make me bullish on this? So one thing is, at least in the context of United States, the cost of labor is one thing that impacts agriculture. Second is, at least from the context of Iowa, the average age of a farmer is becoming higher and higher and higher. The farmers are older and older. And so having AI enabled autonomy is significant. Third is increasing cost of chemicals. And so having AI technologies enable precision farming, right? So when to spray, where to spray, how much to spray, can essentially transform the bottom line of farmers across the globe. And then fourth one is in terms of uncertainties, uncertainties in weather, climatic conditions, uncertainties in trade, uncertainties because of weather changes, the amounts of pests, insects and diseases coming in. These are all significant uncertainties. And AI is poised to potentially mitigate some of these uncertainties. And that's where I think AI is going to revolutionize, if it has not already, how farmers are going to feed the world.
- [08:28]
B
BHASKAR Ganapati Subrahmanian Reducing fertilizer and weed killer use isn't just good for farmers finances. It could help the environment too. This is the sound of Tama tomatoes being mechanically harvested in northern Italy for Muti, a global brand. Tomato plants require lots of water and the Farms that supply the company often overdo it to avoid losing crops in the Italian summer heat. Mutti is now using AI sensors to make sure plants get exactly the amount of water they need and not a drop more. Francesco Mutti is the fourth generation of his family to run the business.
- [09:12]
G
Biorestore is a sensor that is placed on the plant to consider the needs of water for the plant itself. So let's consider that 70% of the water utilized on the earth is for agriculture consumption. If we are able to measure exactly what are the needs of the plants and the terms of water that can reduce on the tomato field. The consumption of approximately 45%, which is a huge amount, definitely cannot be done just asking farmers to reduce the water. Because if there is a lack of water, farmer will risk the entire crop having this sensor that is analyzing help in reducing this quantity. Just if I can give you some figures related to the cultivation of tomatoes. Usually 50, 60 liters of water are required to produce a kilogram of tomatoes.
- [10:19]
B
So you're saying then that you are saving water, your suppliers are saving water because of this new AI technology.
- [10:25]
G
We are studying the possibility of reducing the consumption. In this moment is not for fully installed. That means that one thing is to give the study and what can be achieved. And then you have to train the farmers and to split it among them so that the culture start to become more and more developed.
- [10:51]
B
Artificial intelligence is very energy hungry. And if it isn't powered by a green energy system, emits carbon and adds to the gases that cause climate change. But you're not concerned about that.
- [11:04]
G
In the Biostore project, the solar energy gives the power to measure the plan. Because in that case it's really must. It cannot be attached to a normal source of energy. Has to be the sun.
- [11:23]
B
Does part of you think you're moving away from the the tradition of the tomato growing and processing that your ancestors started?
- [11:31]
G
My ancestor made some very important innovation in this sector. Just two words. One is the creation of the tube in 1951 in Mutti. So the tomato pasting tube is born in Mutti. So innovation fits perfectly with the intent of developing the best quality of the tomatoes. We have to be really careful in maintaining what are the pillars like the quality. But constantly researching what are the best innovations to ameliorate our quality, our way of working, of our workers and in general the health of the planet and of the company.
- [12:10]
B
Francesco Mutti, you're listening to Business Daily from the BBC World Service with me, Rob Young. Today we're examining the cutting Edge of farming, the use of artificial intelligence in agriculture. So how far could AI take farming?
- [12:28]
H
Welcome to a new dawn in farming.
- [12:33]
G
John Deere autonomous tractors Precision power Possibility.
- [12:39]
B
Jamie Hindman is the chief technology officer for John Deere, the world's biggest agricultural machinery maker. So will farmers soon be able to leave the work entirely to autonomous machines?
- [12:53]
H
You know, our intent with developing these solutions, in particular the autonomy solution as an example, is to provide a farmer another tool in the toolbox. This is the way I think about it, where they have the choice of using that tool or not. We still leave the operator station on the machine. It's still on the tractor in this case, because there are a great number of jobs that are done on the farm that still require a human to do. What our intent is, is to give the customer, give the farmer a choice to determine what jobs they want to do themselves moving forward and what jobs they want the machine to do moving forward based upon their individual business conditions. I think we're a long ways from being able to contemplate this world where the farmer doesn't need to farm anymore. I think it's core to who a farmer is. Speaking as somebody that's only a few generations removed from a farm, it's core to who a farmer is that they want to be able to do some of this work themselves. And I think that's an important aspect that we have to keep in mind. But we also want them to be able to meet the challenges of the current time. And if they're challenged with labor availability, they should have a tool in their tool set necessary to address that challenge.
- [14:00]
B
Jamie Hindman from John Deere so far we've talked about how AI is being used in a handful of rich nations. So what about farmers in emerging economies? Cutting edge technology from the US And Europe is expensive, but small scale farmers are accessing some AI tools too.
- [14:19]
E
I am Esther Kimani, the founder and CEO of pharma Lifeline Technologies. I grew up in a rural village here in Kenya called Nyandarwa. And I got to see farming. This is where I joined my family, my mom and dad on the farm. And I remember planting seeds even with my dad in different seasons. And it was very unfortunate that I got to see the devastating effects of pests and diseases where at least a third of our produce was destroyed by the pests and diseases. And this affected our yields. Hence our income was affected. Some of my basic needs, even as a girl went and met and I was actually the only girl from my village who made it to the University. I was just fortunate enough to make it to the university and study computer science. I learned a lot machine learning and towards the end a little bit about AI. And I think that's where I had like this moment of realizing what I'm studying in school is actually presenting an opportunity to solve this pressing problem back home of pests and diseases.
- [15:31]
B
So tell me what technology you have developed then and how it works.
- [15:36]
E
I founded Farmer Lifeline in 2019. Farmer Lifeline exists to get farmers ahead of of crop pests and diseases. And how we do this, we use solar powered AI enabled cameras that detect pests and diseases on the farm and notify the farmer through a simple SMS on their mobile phone. And this SMS will contain the crop that has been planted by this farmer, the pest or disease that has been detected, and recommendations of their most affordable and environmental friendly solutions that they should apply and the exact quantities.
- [16:15]
B
How does it detect pests or disease?
- [16:18]
E
Machine learning is training a machine, or rather this device, to see healthy and infested crops many times to an accuracy that we have right now of 97% until when it's now placed in the real world scenario where it's seen vegetables, fruits and the like. It's able to detect a pesterer disease within the first seconds of infestation. So there's that part of the learning of the machine, then the AI is able to distinguish the specific pest, the specific disease and then now run our algorithms for the exact recommendation that the farmer should use.
- [16:55]
B
What's the cost of it then? How much do farmers pay for this and how does that compare to the increasing yields they're seeing as a result of having access to this information about pests and diseases.
- [17:09]
E
Only $3 per month for them to receive this SMS on time and recommendations of what they should apply. So when they use the technology, they are assured of a 45% increase in their yields. At least it means that there's more money in a farmer's pocket at the end of the harvest because most of the crops were not destroyed and they're not able to sell more food for them to sustain their families. And two, they are also not using too much farm inputs in terms of pesticides on the farm, hence saving them money at the end of the day.
- [17:44]
B
Esther Kimani there, so far so optimistic. But every new technology creates winners and losers. Mehtadavari is from the International Institute for Tropical Agriculture which works to improve livelihoods and enhance food security in poorer countries.
- [18:02]
D
I think there's no doubt that AI holds great promise to increase both yields and profits for farmers. But you know, as in any sort of data driven endeavor, you've heard of garbage in, garbage out. And this holds even more true for AI. Machine learning has been used for years in agriculture to optimize fertilizer application and profit associated with that, to model pest and disease probabilities and provide earning early warning in weather predictions. So there is a lot of potential. But if these kinds of solutions are fed data that is poor quality, that's inaccurate, not representative, it's completely agnostic, it will still provide answers, but they may be highly erroneous or hallucinatory. And since AI also offers a potential to disseminate information really quickly, really widely, these hallucinations do increase the potential for harm. So we do need to be careful with that. And I'm not really sure that we're taking the downsides into account sufficiently. It seems that there's a rush to embracing AI as sort of the solution, you know, to the challenges of agriculture, from the scientists to the funders to the, to the developers of the AI.
- [19:14]
B
Meda Davari. As we heard earlier from California farmer Daniel Alameda, some of the innovations are so new it is hard to know if they will pay off. Thank you for listening to this edition of Business Daily with me, Rob Young. You can get more episodes on the BBC website@BBC.com and wherever you usually get your downloads.