One Farmer’s Thoughts on the Real-World Practicality of Implementing Gaming-Inspired Remote Controlled and Autonomous Equipment
Posted On December 9, 2021 by Stephen Douglass
Part 1: Real-World Farming
A Quick Introduction
After a brief Twitter message exchange with one of my all-time intellectual heroes, Balaji Srivinasan, he being a polymathic entrepreneur and previously a Stanford lecturer, and I being someone just (barely) smart enough to be enthralled by his projects and teachings, I have endeavored to put down here a response to his question to me “how close is Farming Simulator to your actual job?” And by inference, can farming operations be controlled virtually using this type of technology?
For those like me who weren’t familiar with the video game, Farming Simulator, it is shockingly realistic. Not only are the graphics and user interface beautiful and life-like, but the tasks and workflow are far more real-world than I expected. The website’s tutorials look like a proper Ag academic curriculum. The “mods” page is a collection of tractors, loaders, trailers, tools, implements, fertilizer and pesticide inputs, and even buildings created by a dedicated community of coders. If you go to the “live” tab of the Farming Simulator Youtube page here, you can watch people from all over the world playing in real time. It’s fascinating, complex, and realistic even down to market considerations and profitability.
As I watched the game, I found myself wishing I could operate my own farm similarly, i.e. from my office. Keyboard commanding equipment resources, especially ones that can operate semi-autonomously once engaged, would be a dream come true. So, to answer Balaji’s question, yes, it can be done, but there will be a steep learning curve as the myriad environmental variables are solved-for in the programming. Let me explain, but first a little primer.
To be a farmer is to be a Renaissance man/woman by necessity.
As someone who toggles back and forth between the blue collar, hands-in-the-dirt world and the professional class office world, sometimes several times a day, I can tell you that most non-farmers have inumerable misapprehensions about farmers and farming. You have to be at least passably skilled as a mechanic, welder, truck driver, equipment operator, meteorolgist, botonist, agronomist, accountant, chemist, surveyor, builder, carpenter and commodity trader. And that’s just to start. You routinely handle myriad pesticides, fertilizers, chemicals, fuels, solvents, and sometimes literal shit in their most concentrated forms. Farmer’s know and use geometry and algebra daily and usually in their heads to calculate everything from greenhouse layouts to liquid fertilizer and pesticide solutions and other inputs. A farmer firend of mine’s home-schooled son, who signed up for an advanced caluculus class freshman year at University, was warned by the professor that his education was “insufficient for this level class.” What the professor did not know is that the father and son routinely interrupted dinner (the dining room doubled as classroom) to jump up to the chalkboard and debate differential equations. The son’s pre-curve grade was a 103. No other student broke 50. Don’t underestimate farmers’ ability to adapt to new technology.
Farm Tech is already much more advanced than most people realize.
While I spend a lot of time working with and for row-crop farmers, my own farm focuses on specialty produce. Creating a crop plan that accounts for plant family rotations and successions planned through time and space takes a lot of knowledge and effort. Most produce farmers I know create huge multi-booked spreadsheets but there are some fantastic web applications like Tend.ag that have excellent user interfaces for both computer and smart phones. These applications can assign tasks to workers on their own phones, estimate revenues, and merge with the farmer’s preferred accounting application.
Commodity crop farmers use enormous GPS guided sprayers that adjust the concentration of inputs on the fly according to GPS-tagged soil sampling and geographical soil type mapping that has been databased. These targeted and specifically measured applications save money and help farmers put out the “minimum effective dose” which is good both fiscally and environmentally. Long gone are the days of spreading 10-10-10 and the latest greatest pesticide over and over whether needed or not.
The technology that goes into modern farm machines is impressive. The cab of a modern harvester is a comfortable mobile office that drives itself half the time, literally. I know several row-crop farmers who spend their whole day studying commodity markets while simultaneously picking corn. One mistake that machine manufacturers, John Deere in particular, have made recently is to create operating software that prevents the owner from making repairs himself. Imagine a self-reliant farmer driving his half-million dollar harvester on the last day of the season. The rains are coming but he has plenty of time to finish the season with these last 50 acres. Suddenly the Harvester shuts itself down and alerts the nearest John Deere service center (50 miles away) to send a technician to perform scheduled maintenance. The onboard computer prevents the harvester from restarting and the service department won’t tell the owner how to reset the computer because it’s against his service agreement. * The farmer climbs down from the harvester, walks a mile to his truck, checks to make sure his shotgun in the gun rack is loaded, and heads out on a 50 mile trip to the dealership. Kidding aside, farmers will reject anything they can’t fix and maintain themselves.
*This story is a fictionalized anectdote based on real-world lawsuits filed by farmers for the “right to repair.” In our neck of the woods, Deere equipment is practically a status symbol, but the right-to-repair issue has causedused Deere equipment resale values to rise dramatically. And to their credit, Deere tractors last forever.
The equipment used in all farming is much more expensive than people know.
For “row-croppers” who grow and harvest wheat, corn, beans and sometimes cotton, a standard harvester USED starts at $325k. Whenever you get stuck behind one of these behemoths on a rural road, just keep in mind that unless there’s an Italian supercar behind you, that harvester is the most expensive vehicle on that road. That’s the AgroLambo you’re desperately trying to get around. And by the way, DON’T DO THAT. I do almost all the truck driving (transferring grain to granary) for my neighbor row-cropper, and a significant portion of my time behind the wheel is spent preventing new-rurals (aka urban escapees) from killing themselves trying to get around me on my left – regardless of yellow lines, on-coming cars, or even the fact that I’m slowing down 80,000 pounds of truck to make a left turn.
Farming is dangerous as all hell.
I myself have gotten the helicopter ride out of the field in a near fatal accident after a log fell off of a loader, splitting my head open and giving me a sub-arachnoid hemorrhage. I’m better now but I’m told that I have spells where asl;duijhosal;thfjpoiwkrgjampoal;s’dfo,[l;sdgf. What was I saying? Oh, right, danger. Yes, farming is ranked #7 most dangerous occupation, but ironically, of the top ten all but fishing and flying are jobs that farmers often do for themselves anyway. Come to think of it, our preferred crop-duster who, himself, was a pecan farmer, died just recently in a crash. This video is from years before.
Farmers are a conservative bunch, but they are generally open to new methods.
While it’s true that farmers will stick to tried and true methods, a lot of them consider themselves to be entrepreneurs and they are always looking to improve methods, materials, and equipment. In the case of row-crop farming, they are at least somewhat beholden to the input suppliers and the market makers and so don’t have as much control as they want. This is especially true of poultry and hog farmers whose entire operation is married to an industrial level corporate system that is their banker, supplier, and buyer, and, who invariably changes the rules (e.g. food safety) by regulatory capture just when the farmer comes near to autonomy. It’s for these reasons that I think farming can be greatly improved by using a disruptive technology (in the true Clayton Christensen sense.) The question is will that new tech obviate the farmer altogether? Or, will the farmer use that technology to become less dependent on Big Ag while simultaneously switching to a safer, higher level of work where he’s coordinating machine activity instead of being just another part of the machine? This question is worth revisiting later on as it might be the most important consideration.
Part 2: How Gaming-Inspired Farming Applications Can Be Implemented
The farmer will have to be the remote operator.
There is no way the land and machine owner is going to allow someone in a remote location, whose only experience is virtual, to operate his equipment. Therefore, the application must have an intuitive UI/UX. Farming Simulator 19 has already solved that problem. Anyone can learn to use it. Obviously, one could be trained by gaming in simulator mode, and then the next step would be real-world skills – taking off the safety, so to speak.
The remote operator will need to be onsite.
For the foreseeable future, the farmer/remote operator will have to be onsite, or have a designate, to make real-world repairs or adjustments as the system is improving. Lex Fridman, a podcaster and A.I. engineer, recently talked about the fact that autonomous vehicles like trucks will most likely start out as a follower to a human driver. This makes perfect sense. I can see where a harvester may have a remotely operated, semi-autonomous secondary vehicle (e.g. a grain cart) that takes the load from the harvester while it’s underway. This keeps the largest, most important machine operating at it’s highest value.
Farming Simulator will work better for the biggest farmers, at least at first.
The largest operations, like Welker Farms of Montana, will most likely take the lead in automation and remote control. They already have a Farming Simulator 19 Mod and Map where you can tour the farm itself virtually. There are several videos of Welker Farms game play. The obvious advantage here is the fact that these mega-farms are self contained, meaning they have all of their own equipment, refueling and repair facilities, and post-harvest storage. I wouldn’t be surprised if some of them have a spur track for hooking up railroad cars. On a large, self-contained farm, it will be much easier to standardize fields and processes, and also minimize dangers to humans. Smaller East Coast farms are typically split up geographically, and requite a lot of transport on public roads. The operation usually involve a lot of leased land as well, which would make securing them for human safety much more difficult. Remember, these machines chew up all the fauna in their paths so safety will be the highest priority.
Part 3: Conclusion & The Larger Question
As an old, beat up farmer myself, the idea of operating the equipment remotely is incredibly appealing. It’s a job promotion when you think about it. It’s an opportunity to level-up and to become the conductor of the farm symphony, and maybe to be perceived as more professional class? Not to mention, it will make farming less labor intensive, more efficient, and more profitable.
Here’s the rub, though. Since this has to be implemented first on industrial scaled farms, will it further cement a farming system that could very well be responsible for an obesity epidemic and other health problems, as well as environmental degradation?
In my opinion, the COVID-19 pandemic has exposed the fragility of our food system; a fragilty caused by centralization. Check the two earlier posts on this site on the subject of “robustifying the food system.” The first one is more entertaining (as it involves Joel Salatin,) but the second one is worthy, too.
Perhaps the makers of Farming Simulator 19 can gamify a small farm platform based on J.M Fortier’s highly standardized and successful produce farming system. If so, we could have tens of thousands of decentralized, ecologically sound, financially successful, local farmers securing the food system like validator nodes of a blockchain. For that matter, why don’t we create a network, a community, and a currency for just this purpose? HA! Don’t get me started on that! Hmmmm……maybe next post. 🙂 ——Farmer S