Yup. A few days ago we (60 people) got bought by John Deere for $305M b/c our machine reduced herbicides by 90% using CV/ML. Our only moving part was a solenoid. Would not have survived if it had been more complicated. You clearly get it.
Blue River Technology | Sunnyvale, CA or SF, CA | Onsite | Full Time
Agricultural Robotics. Hiring all types of SW Engineers at all levels especially those with the willingness and aptitude to delve into machine learning. We made a machine that reduces herbicide usage by 95% and dramatically lessens the necessity for GMO's.
Write SW for machines that are 50 feet wide and learn to drive a huge tractor while you're doing it. You'll save farmers a lot of money and keep the rivers blue while you're doing it.
Thank you. I work in Ag and it seems crazy that someone should be able to buy a machine like those Deere sells and be able to modify the controls SW. Can United Airlines update the controls SW on Boeing 747? I'm not sure how tractors are any different.
I also work in ag, and own John Deere equipment. Is anyone truly asking for source code? I've read that narrative here and there, but it seems what is really being sought after is access to the tools and manuals that the authorized dealers have access to. The service techs repairing these machines aren't going to be modifying the code either, and they don't need to.
As an aside, I remember the last time I had a John Deere tech out. Even for him to access the service manual (to fix a mechanical part), which was fundamentally not much more than a simple search tool that displays PDFs, required a surprising level of authorization to access each document, with everything encrypted. It seemed a little extreme just to get a simple diagram. I can sort of see why some worry about where the company is headed.
I think documentation is the right way to handle this. Require enough documentation that 3rd party repair shops can fix the machines. Even if they can't do much with the embedded controllers, at least document the sensors and the wiring. Otherwise, these machines are completely useless without the manufacturer and their dealers.
Can confirm that United Airlines directly employs aircraft mechanics who are free to service as they see fit, and they use hardware that is not DRM'd to the plane.
Can also confirm that aircrafts can do more damage than tractors if things go wrong.
Like the JDs in the article that are so complex that they can steer themselves, aircrafts also do this!
I don't see how tractors are any different either. So JD should cut out this rent-seeking bullshit and stop pretending it's in the best interest of anybody other than their shareholders.
From all the articles that I have read, the repairs of the United Airlines planes are heavily regulated. The mechanics can't replace a part with another unless the new part is certified, and certification is very expensive. The mechanics also have to be certified and fill a lot of paperwork.
The changes checks are not enforced by software, but by strict FAA rules.
United Airlines is not really farmer Joe and something tells me that even United Airlines doesn't send its 747 to Boeing for all the maintenance work...
Relative to HN I'm probably an expert in ag. Relative to people who grow things I'm far from an expert. I was employee 3 at Blue River Technology and I've been there for four years: http://www.bluerivert.com
The way the Valley generally thinks of ag is completely broken and I highly recommend avoiding Valley people if you want to learn about it. "Fully automated" is a thing people who've never farmed before assume is possible and assume people who farm want.
There is a lot of value on the table in ag tech. Depending on what your objectives are, I recommend having as many face to face / dirty boots meetings w/ growers as you can. Understand the realities and nuance of their day to day and find a place where you can make their lives better. Ignore the sensational articles. Most "ag-tech" companies are comically detached from the realities of their intended customers.
If your goal is to grow things then follow the above advice w/ 10x the emphasis. Happy to help further. Contact in profile.
Impossible may not be the right word, but highly unlikely due to the environment ag automation needs to operate in. Production agriculture is highly mechanical, meaning things wear out and break all the time, which means humans need to go and fix things. In addition, people farm in every environment on earth, with wide variance in conditions in even the most ideal places. This means mechanical equipment needs to be specialized for the crop or location (considering humidity, temperature, wind, sun, crop type, etc), which makes scalability limited. These are the two biggest issues that tech doesn't really have a good solution for.
I'll take a quick stab. First off it's not a software computer system. Their are so many different elements in play. Your robot needs to know about soil science, agronomy, pests, weather predictions, stock markets, machinery, the list goes on.
Farmers have to wear many different hats. You are a seed scientists for part of year when you figure out which seed you should plant for next season. This choice involves many other factors (weather predictions, personal yield goals etc). During planting season you have to make human decisions of when to plant/not plant. Training a robot to make this decision would be very complicated and most likely involve many 'failed' attempts that cost tons of $$$.
Lets say you did get the robot to choose a good seed, then you hired a fleet of robots to plant corn on 5000 acres and they successfully chose the correct times, right plant depths, row spacing, seeding rates, and managed to get the crop fully planted without any hiccups.. At this point you would then need to be scouting the fields, most likely putting down some N after planting. Robots have to decide which fertilizer goes where, how much to put on, is each field the same or does field X get higher N on one side but less N on other??
THEN lets assume all your crops grow fine, weather behaves (never does)and your robot fleet has scouted all fields. Then the robots must go out and harvest the grain. The combine picks up grain which then gets dumped into a semi on side of the road. Again maybe there is a 'harvest semi for Uber' kind of service where automated semis drive around between combine and elevator at will.
Another decision to be made is selling the grain or storing it...
In short, there are way too many decisions and farmers (i'm thinking midwest corn/beans guys) have to make across many disciplines. These guys have an intrinsic knowledge about their land that has been accumulating for 10s or 100s of years!
Growing up on a family farm & surrounded by farmers, most of them are not great at any of that. I do think there is a lot of potential for automation for each one of the things you mentioned. At first it can help farmers make better decisions. Slowly it can grow to make those decisions on its own.
A 'huge' part farmers struggle with is trying to guess when to purchase & sell their products. That's everything from the end product to the seed & fertilizers. Also, tech could help farmers get a heads up on if there might be global surplus or shortage on certain crops. Sensor based tools can do a much better job on how much & where to put fertilizer. I've even seen demos of not using pesticides but friggen laser beams to shoot only certain types of insects.
My prediction is future farms will look like giant JavaScript projects with crazy long package.json files!! There are so many possible tools that could improve a farmer's life. I can even see "Tool Fatigue" being discussed at the local gas & coffee shops.
It doesn't change anything, but as the Uber engineer above lists, they have rational reasons for wanting it.
I just don't have any rational reason to want to give it to them, so I'll be disabling GPS entirely. But I've mostly stopped using them anyway - the scummy corporate behavior bugged me, but not enough to boycott. And then a driver nearly physically attacked me over, as best I can tell, mishearing something (although I'm not really sure). Those combined, along with a general dislike of the gig economy and I've already mostly ditched them - Uber feels somewhere between sketchy and icky to me. Used them once in the last ~6 months, when the Lyft app forgot my password and it was late.
They let me not let them. Lyft lets me tell iOS to share my location with it only "while using the app". Uber used to do this. But now that option has disappeared--it's only "always" or "never".
You don't know until someone investigates, obviously. But if you have to choose between an app that definitely does that and an app that might do it, it makes sense to use the one that might in the meantime.
There's an opportunity for Lyft to be "Uber, but for not scummy business practises" here, but I suppose most users don't really care. Austin would have been a good test of that, but Lyft withdrew along with Uber.
Lyft, sadly, doesn't seem to have the talent it needs to pull this off. Although they're still worth something, it just doesn't seem like Lyft will be able to eke it out unless something really game-changing happens to Uber.
I've been a longtime Lyft user because of the stories that surfaced a few years ago about Uber considering blackmailing reporters and politicians with their ride data, but the last time I had to use the app, Lyft for some reason just wasn't working - their auth service seemed to be broken. I ended up downloading Uber and signing up because I had to get somewhere in a hurry and didn't have the time to look up another rideshare service. Their text notifications about driver arrivals are also always off, on the order of up to a minute or two (I don't know if Uber is better, but I would assume it is.)
FWIW, I'd also sunk something like $300 into Lyft rides just two weeks before that for business travel, and I didn't experience a problem once.
Lyft engineer here, so bear in mind any bias, but we take account and data privacy very seriously, and have a data privacy team which focuses on these kind of issues.
From where I stand, we have great engineering and product talent. Perception is tricky - I see that we're very agile with (sometimes far) fewer resources, from the view I have.
I'd be glad to dig into the auth issues you ran into if you're able to ping privately with a few more details.
On iOS at least - the (current versipon of the) Lytf app specifically asks " … to access you location while you use the app." Uber _used_ to have a "while using the app" location option, but the recent update took this away so you choice is either "Never" give it location info, or "Always".
That parable is misleading. If his window wasn't broken then the shopowner would've invested it in some improvement to the store or as a safety cushion which would also trickle through to the relevant suppliers. We'd actually all be better off if the window wasn't broken because we should have some other investment that was equally valuable without having to lose the asset of the window.
I don't think "Parable of the broken window" captures the essence of our economy, it only mentions the handyman receiving 6 francs, and not him paying his suppliers, them paying their suppliers and so on, his tools, his rent, his taxes, etc.
> I don't think "Parable of the broken window" captures the essence of our economy, it only mentions the handyman receiving 6 francs, and not him paying his suppliers, them paying their suppliers and so on, his tools, his rent, his taxes, etc.
That isn't actually a multiplier. When the handyman is paying other people it means he isn't actually making $6. He receives $6 but then the tax man takes $2, the window costs $1, various other things cost another $2 and the handyman only keeps $1. No matter how you slice it the grand total has to be $6.
But you don't want a multiplier here anyway, because it's $6 of waste. If the $6 somehow turned into $100 then it would be that much worse.
The multiplier effect exists, but in the broken window fallacy all it does is multiply the wasted effort.
The company mining sand is wasting sand and people's time mining the sand for the broken window.
The company that melts the sand into glass is wasting time and energy making the glass for that window.
The company that transports the window to the distributor is wasting storage space, energy(gas), and salary transporting it to the retail location(Home Depot).
The retail company is wasting storage space on the window, and perhaps a couple minutes of a salesperson's time.
The handyman is wasting his time when he installs a window to no benefit.
The insurance company is wasting their time handling the claim, and is implicitly costing everyone else in their client pool time with increased premiums.
And the person whose window was broken has their time wasted working in order to pay for the new window.
Any property that any of these people use is wasted, because if your office processes claims for broken windows the portion of the rent used for office space is wasted. (For example, if you spend enough time on broken windows, you might need extra staff and therefore a larger property.)
In each of these, maybe only a few minutes, seconds, or milliseconds(in the case of the sand needed) is wasted. But the entire ripple effect is ultimately a waste. In some cases, there appears to be no effect because a company may have additional capacity already available. But if you make a complete accounting of all effects, it is all a waste.
When I look at issues in economics, I like to erase the money entirely, since money is often nothing more than an accounting tool to make a graph reduction problem easier. In the case of a broken window, all you've done when you eliminate the money is waste resources, land usage, and many people's time.
Contrast this with a new window for a brand new building: Presumably that building is being used for some productive purpose, so the entire supply chain is implicitly being used for a productive purpose.
The problem is that if you could magically remove all the waste in the economy, but didn't find new work for people to do, then all that happens is you have a lot of people out of work, which is in itself a waste. The way our economy is built, the waste from people being idle is in many ways worse than the waste from people doing things that don't need to be done, which is why politicians are always pro-jobs and never mind whether the work is worth doing.
Leaving aside money, efficiency should never be a problem because there is always plenty of work that urgently needs doing. Freeing up people to work on more important things should always be a win.
That's not the world we live in. To explain why, you need money: nobody will pay for it!
Convincing people to spend money on things that urgently need doing is the heart of the problem. The broken window doesn't have that problem. It's a waste, but it's a very convincing reason to spend money, and the shopkeeper has money, so that's what happens.
Building a new building could be a waste, too. Overinvestment happens and new buildings do stand idle sometimes for lack of tenants.
I agree that the economy exists to serve people. The fact that the economy demands people get jobs in order to be served and then doesn't give them jobs should be seen as a damning case where the economy isn't performing the function for which it exists.
If we had something like basic income then maybe you'd have a point. But spending a lot of time looking for a job and not finding one is certainly a waste. Deferring expenses (like going without medical care) is a waste. And crime causes lots of waste. So, back to broken windows again.
GMO + broadcast applied herbicides is a long term loser. For years cotton growers in the South Eastern US applied Round Up from planes. The chemical would dissipate on its way down and apply half doses to weeds. Some of these weeds could metabolize the Round Up while others would die. But the next year all that was left were the weeds that had some resistance. They bred w/ e/o and on it went. The annual selective pressure has created some pretty impressive weeds, most notably Palmer Amaranth. Growers I've talked to say, "Round up just pisses it off." It can grow over an inch a day, it has over a million seeds in its bud and once it has grown past 4 inches, no selective chemical can kill it. You can kill 98% of it one year and its population will be bigger the next year. Once you get an Amaranth population, you more or less can't get rid of it. It's a one way ratchet. As such, Palmer Amaranth is spreading.
The Dicamba story is quite interesting. Monsanto is selling Dicamba resistant seeds even though the chemical is not approved for use on those crops. As Dicamba is approved for other uses, growers can buy the Dicamba ready seeds and illegally apply it. Dicamba is finicky. In many weather conditions it volatilizes and drifts to a neighboring farm. If the neighbor doesn't have Dicamba ready seeds, he could lose his crop and have very little legal recourse. Today two growers got in a dispute about Dicamba drift and one shot and killed the other. The drift problem is forcing growers that don't use Dicamba to buy Dicamba resistant seeds as a defense mechanism. It's pretty nuts. I feel like the herbicide resistant weed problem is like a "Global Warming that nobody knows about."
More interesting is the fact that it takes decades to produce a gene / chemical pairing and get it approved for use. The weeds evolve faster than the GMO / chemical companies' ability to deliver new products.
The startup I work for, Blue River Technology, is going after this problem. We make machines that detect the crop, detect the weeds and apply non-selective chemicals to just the weeds. Because we target selectively we can use non-systemic, contact herbicides will kill resistant weeds no matter what. The grower saves on chemicals, put orders of magnitude less poison in the ground and gets better weed control.
The problems w/ this article are numerous. The guy clearly spent more time looking at UN stats than he did walking the fields and talking w/ growers. Somehow he missed the rather obvious fact that GMO + roundup is more cost effective regardless of yield. Regardless, he brought up some interesting points and I'd rather be talking about it than not.
-- Disclaimer --
The company I work for is partially funded by both Monsanto and Syngenta. Not that anything I said is defending the status quo :)
The main reason to avoid any sort of mechanical kill is time. Even if you could deploy and retract a puncher in 100 ms, you'd travel 10 inches during that time. The slowest operation in row crops is about 6mph. Same principle applies to microwaves & lasers. All require time over the target. Chemicals are nice b/c they do their work after your machine leaves and its easy to get a solenoid that opens and closes in 5ms.
There are other issues w/ exploding seed head making the problem worse and w/ making reliable complex moving parts in the horrible ag environment.
