John Deere unveils fully autonomous 8R tractor
agriculture.com
agriculture.com
So now your tractor is controlled via their network? If their server goes down, can you still farm? What if your John Deere Operations Center account has been been terminated or suspended? Here are their terms.[1] "In its sole discretion, Deere may restrict the Site or restricted areas of the Site to selected users."
Farmers have been fighting for the right to repair Deere tractors. So this is Deere's response. They now claim the right to prevent tractor owners from using them.
https://www.theverge.com/2020/8/11/21363122/boeing-747s-flop...
Wait, maybe Jeremy Clarkson bought that lambo tractor under his own name. But it is more likely that his production company owns it.
The fact is, a pilot did not purchase the plane with his own money, and is working for a giant corporation that can negotiate a contract with another giant corporation.
A farmer, esp. a small scale private farmer, is not in the same position to negotiate a favourable contract as American Airlines would to Boeing.
It’s a pretty even fight, and because of the existence of Airbus, the airlines actually have a slight advantage. Look at what happened when Southwest casually mentioned they might be thinking about adding A320s into their fleet. Sent Boeing scrambling.
Boeing actually seems to have more diverse competition, including Lockheed, Northrup, etc (although they're dominant in the commercial airplanes segment, where airbus isn't mentioned, so maybe these stats are US market only): https://csimarket.com/stocks/competitionSEG2.php?code=BA
Being able to get service in the field quickly is critical during planting or harvest. During these times downtime can be exceedingly expensive if the farmer is trying to beat some change in weather.
And sure you could try and get someone other than the dealer to work on it, if you can find them. But trying to find this person when your equipment is broken and you are burning money isn’t a great option and the dealer will probably favor servicing clients who bought equipment from them.
John Deere manufactures it’s own engines in many cases, but it also uses engines from other manufacturers.
> never (intentionally) leaves the ground or exceeds 30mph
Yes, aviation has a lot lower risk tolerance than agriculture, but ag equipment still kills plenty of people, and I’d probably rather be in a head-on collision with a honda civic going 50mph than an 8R going 30mph.
> Most medium sized countries have tractor factories, as they're essential for food production for the population.
Are you just reiterating that John Deere has more competitors? If so, I think this is misleading because most of these factories are just the same few competitors with factories in different companies (rather than distinct companies per country).
Flying your modern aircraft doesn't depend on a lousy app that requires your "username and password" and probably doesn't even know what a pwd salt is. And if it goes down you can still fly (or keep flying).
You won't get cut out of replacement parts unless you're in a sanctioned country (and they do have ways of sourcing the parts).
Yes the agribusiness are ok with this, they know how to handle it (but you'll bet they'll have a lot of angry calls and lawsuits in some cases). Small farms on the other hand...
But what happens when JD discovers a problem with the software? How do they force you to upgrade your software? What happens if the bug is dangerous or the bug causes your tractor to have mechanical issues if not upgraded?
Does the farmer want to be responsible for installing the emergency software update?
This whole idea of shipping half-baked and fixing later OTA is one of this age's great consumer scams.
Not that it could not happen... there are various hedge funds that do this already, with mixed success. They generally lease back to a farmer.
John Deere doesn't have any data that the farmer wouldn't have, and they certainly don't have better access to large sums of capital than hedge funds. I can seem them participating in a capital lease agreement for someone doing this sort of operation, but operating a farm is a very different proposition.
Sure they do. They have the data from all the farms that use their service. The farmer only has data from their own farm. Deere can tell who's doing better.
As an example, Zillow had all that house price data, and then blew half a billion dollars because it thought it could use that data to start buying and selling houses itself.
The tractor and combines are sending real time crop data, it isn't hard for someone to look at that data and see across all farmers that yields are a bit different from what the market is predicting and make a large profit trading on that insider knowledge.
Those who work on that software and the admins of the database are registered with the SEC as not allowed to trade commodities. All that money they could make is directly out of our customers pockets and they wouldn't let us have that data if were were not protecting it.
I have family that still use JD tractors from 60s and 70s. In fact I have a relative who bough one for less than 2k, refurbished it for ~5k, and use it during winter for his chalet. Why don't we build pre-computerized tractors? I understand that cast-iron will break, but still the point remains with CNC machines becoming cheaper and cheaper.
Because unlike regular passenger cars the automation in tractors is essential for high yields. A farm that uses high tech tractors is going to beat out a farm with traditional tractors every single time. The days of the old boomer farmer are long gone; anyone who has managed to avoid getting swallowed by the conglomerates by this point has a pretty cutting edge setup that would make some startups blush. Drones, IoT, ML, etc. have been penetrating agriculture since the 2010s.
I think a better strategy would be to build open source tractor software for open source tractor hardware; some kind of "Linux for farms" coalition. It is very hard finding software people who intersect with agriculture people though which is what makes this a difficult undertaking.
What would open source bring? You always need a hardware anyway, and the conditions are rough, a RPI would not survive.
And for farmers locked in JD awful customer service, well they have plenty of choices beside JD, except they are not all American.
I would say many software services moving to cloud and to a subscription has largely prevented those products from being pirated or even necessarily owned. Inspired by that, for JD the most convenient path to create an "anti-piratable" product is by slapping a label of "autonomous and AI" and have them being directly controlled from a central location.
I’m plenty critical of John Deere (having worked there for a decade), but I think you have it backwards. Technology drives profit for John Deere, and John Deere’s right to repair is a defensive play to keep third parties from eating into those profits.
Parenthetically, I also strongly suspect that the tech media are overly-cynical about its portrait of Deere’s anti-right-to-repair stance. Practically speaking, it becomes a lot more expensive to ship software updates when you can’t guarantee anything about the state of the system you’re shipping to because the customer has modified various things. Moreover, the more people tamper with these complex, dynamic systems, the more likely there are to be issues, which Deere probably fears will erode its “quality” reputation (Deere’s management can’t maintain Deere’s reputation; it certainly doesn’t want other parties accelerating the erosion). Furthermore, triaging issues and litigating responsibility becomes dramatically more expensive (note that unlike other parts of the software industry, “issues” here are tens of thousands of dollars per incident in the best case and “loss of life lawsuits” in the worst case). So yeah, all of these things are about “profit” from Deere’s perspective, but it seems to me to be less about avoiding competition per se and more about the substantial cost burden on Deere to support a third party ecosystem.
As another parenthetical, I also think people attribute more competence to Deere’s software business than I’ve personally observed. I would honestly love to work on Deere’s technical problems, but Deere has so many political, cultural, and managerial problems that impede technical progress.
Same story with ASML, Siemens and other industrial tools companies.
Things get interesting when the cost (economic value) of the tool is a largest percentage of the total $ of goods produced.
In the 20th century that phrase was written about workers, as in employees, but if even the business owners don't really own the means of production... are they really business owners ?
It seems history repeats itself, just one echelon higher in the society.
Or did JD increase farming productivity in a pareto efficient way, so the farmers are still better off buying JD, even though JD gets to keep most of the newly created surplus?
Or is it just a race to the bottom where farmers that don't buy JD get priced out, even though buying JD makes the farmers worse off as a collective?
Tom Philpott describes this really well in 'Perilous Bounty': there are lots of farmers who are price takers. Their only way forward is to hope to survive weather and price cycles by producing more, more efficiently. This ironically leads to overproduction and poor returns.
Whereas JD are price makers who can use their consolidated power to set terms and prices within the market. Farmers don't really stand a chance.
But it is certainly true that production subsidies serve to make landlords and suppliers rich just as they serve to make farmers rich - your tax dollars at work!
My grandpa probably has a few million dollars in assets, but he saved aggressively his whole life and probably worked 40% more than the average non-farmer over the same time span (longer hours, fewer/shorter vacations, retired in his late 70s, etc).
Yes, farmers have a lot of capital invested in their businesses, but their incomes (i.e., what they take home rather than reinvesting in their businesses) still seem pretty modest.
Compared to my cousins who farm, I’m pretty sure I’m wealthier than they are as a software engineer, certainly with respect to income and perhaps also with respect to total asset value.
It’s bound to happen eventually, but we probably haven’t crossed that threshold yet.
Grain farming doesn't become a full time job until you reach over 1000 acres, if a guy farms one section of land and makes a decent living and chooses to spend that extra time enjoying life, more power to 'em, but farming in that 640-1280 acre pocket scales really well - it is a choice!
The over half of farmers in Iowa, most of which are row-crop farmers, that are running farms under 640 acres, are they dropping out because of economic pressure? Maybe some, but I think mostly it is old-timers that had a great run but they have a business that provides half an income to the next generation at the current scale and the next generation would rather do one thing than two. Either way, Dad's not retiring "poor"!
(This happens a lot in my neck of the woods, lots of 1,000 acre-ish farmers retiring with kids that are professionals that "love the farm" but have no interest in it at this scale.)
The fact that many farmers choose to operate their business in a way that is under-leveraged and often over-work themselves instead of hiring outside help that they can certainly afford is an artifact of culture, not a measure of wealth.
Of course, coming from a rural community there is a pretty good chance that you engage in an unusually high level of personal savings and as such have converted a high income into high asset holdings, I know I do - it is something in the water I think!
Modern sharecroppers.
Usury, rent seeking, monopsony, wealth transfer.
Institutionalized power imbalances which eventually lead to abrupt disruptions. aka Revolutions.
This story is as old as history.
I've been reading more about the Progressive Era during the apocalypse. (Currently chewing thru Caro's Master of the Senate, which covers New Deal thru Great Society.) Same challenges, rhetoric, social fault lines. Only the cast of characters have changed.
Don't worry, though. The more I learn, the more optimistic I become. We've been here before. Countless times. And somehow, humanity stumbles forward. Though the stakes are quite high for this iteration.
If there's anything "exceptional" about American Exceptionalism, my nomination is that our transfers of power have been relatively more peaceful than most.
I'm not saying things will necessarily work out. Only that it's possible we'll reach the next equilibrium. Kick the can down the road a few more decades. Potentially giving our grandkids the opportunity to revisit this never-ending argument again, on their own terms.
Right now, just knowing it's possible is hope enough for me.
Any farming operation where humans might possibly be around this won't work for. We want this to be full autonomous, but the technology isn't ready.
The primary purpose of this is to target great engineers (ie the people who read hacker news) with the idea that we are an exciting company to work for. I personally have some code in this tractor, and I know the people who work on it. We have lots of open positions, the jobs are fun (as hard as any other engineering of course), and you might want to join us.
Everything else can have sensors on board, cars do this.
Suppose the whole Keurig's DRM idea is applied to automated farm equipment. Want to plant some seeds? They have to be from a given source. Want to plant a certain crop, you can't plant it unless it's treated with a given pesticide. Inline sensors detect exactly what you are dispensing.
This all works because if you want to automate the driving, you have to integrate with the dispensing of things: seeds, fertilizers, pesticides, etc. If you don't, you need a person
Of course there is a flipside where that can be used to prevent you from using illegal pesticides, as those same in-line sensors could prevent dispensing of undesired things.
I do know the obvious answers to this but what are the externalized costs of not doing it, and are the trade-off worth it?
Like....some type of legislative system...where they hold hearings and make laws about them?
That would be sick. Some startup should definitely disrupt this space. They could fund it by charging yearly subscriptions, maybe collect them every April or something.
(Jokes aside: this is supposed to be the job of regulatory agencies, but because industry has grown so powerful and been unchecked for so long it doesn't really feel like anyone has any teeth left to stop corporations like JD anymore).
I am imaging the horrors of an HP printer ecosystem on a farm... Which now makes me wonder why HP printers do not require special HP paper.
Fortunately, there is not much moat here. If John Deere demonstrates that self-driving tractors are a massive profit win for farmers, you'll see 10 competitors pop out of the startup sphere within a year.
in the case of cartridges, the Hé cartridges came first, then the compatible ones. In the case of paper, it existed before HP printers.
See also Keurig capsules and water.
IIRC this was the major source of Monsanto's controversies, they claimed that their seed genomes were covered under patent law and farmers were not allowed to grow more plants using the seeds from their existing plants. They'd have to buy more seeds from Monsanto every season.
[1] https://www.prnewswire.com/news-releases/some-of-the-major-c...
Agriculture is a critical component of a functional society. It's not a good idea to allow a few corporations to monopolize our entire food supply, and small farmers are constantly struggling to keep their heads above water. When suppliers like JD pull this BS with tech, it's a canary in the coal mine moment signaling that whoever controls JD could one day indirectly control US agriculture, which is a significant national security concern.
This particular tractor seems to be completely linked to JD's cloud system, there's no way to run it locally or independently from JD. If this becomes the norm 10 years from now and every tractor relies on JD's permission to operate, what happens when JD gets hit by ransomware or attacked by hostile nation states? Every tractor in the country suddenly refuses to work and we no longer have enough food to feed everyone. This would be a manmade disaster with unimaginable consequences (although one easy to imagine one is nuclear holocaust).
You can not image a scenario where this can cause huge supply chain issue with the food supply?
You can not image a scenario where this leads to further consolidation of farms into larger and larger corporations that control a larger percent of the national food supply?
I am shocked more people are not upset about this.
Half of the food grown is thrown out instead of eaten.
Ploughing in your unsellable fruit is said to be heartbreaking. Suicide in farmers is high.
Growing less and getting a better price for it might be better than growing more but losing half the produce to skin taint or bruising. Less inputs, more value per acre worked.
"There is a big shortfall between the amount of food we produce today and the amount needed to feed everyone in 2050. There will be nearly 10 billion people on Earth by 2050—about 3 billion more mouths to feed than there were in 2010. As incomes rise, people will increasingly consume more resource-intensive, animal-based foods."
[0] https://www.wri.org/insights/how-sustainably-feed-10-billion...
That's about the only thing that comes to mind when I hear "John Deere" now.
You can start renting the tractors in the South and keep moving them North. While a farmer might use that tractor a maximum of three months a year Deere could keep them running more than double that usage.
It's possible that a startup could start doing it proving the model and then get acquired by Deere.
Their business model is they buy crops in the field or charge to do custom harvesting. Usually the former.
I've not been onsite for one outfit but the other is close to home so I visit from time to time. The amount of equipment they have is amazing. Millions of dollars of very large equipment. They load it up and truck it around between multiple states to harvest.
Point being, sometimes farmers have little or minimal equipment. They might even contract planting, spraying etc. and sell crops directly in the field. Or they might plant and cultivate but not harvest.
That certainly makes this kind of thing more affordable as it ultimately spreads the cost between many farmers (with the company that owns the equipment taking a cut of course).
This could shift the skills needed to farm more towards automation/programming tech etc
We did get the increase in production efficiency but that newly-generated value was never passed on to those who did the actual work (all the way from farmers and factory works to most white collar jobs). It was captured and collected by those at the top. So now, everyone's afraid of "losing their jobs to machines" and somewhat rightly so. Instead of building machines to do our bidding while we live like royalty, we still have insane amounts of poverty despite the world generating more products/food/value than ever before in human history, with people losing their jobs and getting nothing in return.
It's likely also a leading cause of this: https://en.wikipedia.org/wiki/World_population#/media/File:W...
That's why the wealthy don't hold much cash. (unless they see a crash coming)
https://agamerica.com/blog/the-impact-of-the-farm-labor-shor...
How U.S. Farmers are Coping with the Labor Shortage Crisis
While the USDA is working on new legislation to address these impacts, U.S farmers are turning to alternative options to compensate for their current labor shortage predicament.
Smart Technology Advancements
The Internet of Things (IoT) industry has exploded as farmers use innovative technology to keep up with demand and remain competitive in the global market. The American Farm Bureau Federation (AFBF) reported 56 percent of U.S. farms have begun using agritech, with more than half stating labor shortage as the reason.
Innovative agrotechnology includes,
Wireless sensors; Predictive forecasting models; Advanced automation; Robotics; and Data analytics
Smart farming technology allows farmers to work at full or near-full capacity with fewer workers, but not without certain drawbacks.
The current technology can only accommodate certain crops, excluding nearly 200 specialty crops. This forces specialty farmers to make the difficult decision to either consolidate their operation or transition to machine-serviceable crops.
Transitioning to smart farming technology also requires a high initial cost investment. This is particularly difficult for small to medium-size farms, which are the ones hit hardest by the effects of farm labor shortage.
Despite these drawbacks, many U.S. farmers are realizing the investment is worth the initial cost to enhance crop quality and quantity through precision automation, pest control, and crop-health monitoring.
Farther south, several crops per year can be taken, possibly of different cultures on the same land.
The tractor can work 24/7. This is very important during certain seasons when the windows of beneficial weather are small, and the effect of the conditions at that day on the whole crop is outsized. I suspect tractor operators are in short supply during such season.
I can see how a robotic tractor may feel more dependable than skilled seasonal labor.
https://www.futurefarming.com/machinery/john-deere-acquires-...
https://agfundernews.com/breaking-exclusive-john-deere-acqui...
Don’t get me wrong, the worst case scenario is bad. but anyone interested in creating that scenario probably has better tools at their disposal.
If this tractor gets hacked, it will be to allow 3rd part spare parts.
But probably not to drive places randomly and smash into things. The better attack is to change some engine settings so the entire fleet is ruined just before a big harvest, so there's no time to fix them. Attacking people with tractors just draws attention to your compromise.
https://www.ishn.com/articles/112748-top-25-most-dangerous-j...
The leading cause is transportation accidents. Farmers driving tractors and equipment and workers traveling to worksites.
Actually a lot of the jobs list transportation incidents as the reason.
It's also much less formidable landscape than driving autonomous vehicles among the humans.
1) The area that these machines work is already considered an insulated work space, you would not expect to have humans on the grounds when the autonomous machines operate.
2) These machines make a lot of noise and move very slowly, which would allow humans to protect themselves from machines that misbehave
3) The farming equipment is already bulky and doesn't need considerations for things like aerodynamics, which means that all kind of sensor and extra equipment can be attached.
4) The business model can tolerate cost increase, it's very common for these machines to be time shared between users.
I'm looking forward for a future where being a farmer means machine management from a comfortable room. The traditional farming is so dirty and hard job that most young people are avoiding it at all cost and this is not a business that can cease to exist as farming is what makes our civilisation exist.
Is there any reason why they cannot be geared as required to plough at 50/60/70mph etc? Is there something about the actual physics of ploughing/whatever/etc that requires slowness? Or is it more the fact that it is physically very arduous for the machine and it would cost too much for both the power required and speed?
Just kinda thinking along the Elon Musk line of thought around robots in a car factory - i.e. why aren't they moving so fast that all we see is a blur etc
Another consideration is that a 70 MPH tractor would stir up way too much dust. Topsoil would end up everywhere except where it's wanted.
One valid question might be why a robotic tractor needs to be shaped like a human-driven tractor, or work like one. Maybe it'd be more effective to use an army of small drone-like machines that operate closer to the ground?