Robot Crop Pickers Means Fewer Farm Workers | Hacker News Reader

Robot Crop Pickers Means Fewer Farm Workers

bloomberg.com

44 pointshourislate58 comments

The apple-picking machine from Phil Brown Welding shown not only isn't robotic, it doesn't pick apples. It just supports human apple pickers by putting them in a good position for picking, and collecting the fruit.[1]

Real robotic apple-picking machines are still experimental.[2] Too slow, too fragile, too much tech support required. The video of the SRI International spinoff, Abundant Robotics, shows too many people involved, and too much duct tape. FFI Robotics in Israel has a similar demo system. Someone should build a good one within the next few years. But they're not here yet.

In practice, many apples are harvested with tree shakers.[3] Simple, cheap, and fast, but a bit hard on the tree. Then they're picked up off the ground with another large, but dumb, machine.[4] There's a variation on this where the tree shaker snaps a net around the tree and catches the fruit as it falls.

Once out of the fields, all further processing can be mechanized, and usually is in the developed world. Incredibly fast computer vision systems inspect fruit, berries, and sometimes even kernels of corn and grains of rice. Here's a typical grape-sorting machine, with good slow-motion closeups of the separation of perfect grapes from defective ones.[5]

[1] https://www.youtube.com/watch?v=KJEtz8ejPvw [2] https://www.youtube.com/watch?v=TBcWZcjXr-I [3] https://www.youtube.com/watch?v=Q6kBvc92KB0 [4] https://www.youtube.com/watch?v=lzQ0pfbIW0w [5] https://www.youtube.com/watch?v=vbSww5SBqN4

It's funny how two people can read the same article and come away with opposite impressions. I didn't see anything in the article mourning the loss of these jobs. I won't mourn the displaced jobs either. They're physically demanding, don't offer steady employment, and don't pay very well.

I didn't see anything like Luddism expressed or implied in the article. Is it Luddism if articles about labor saving technology don't always tack on a fairy tale ending that goes "...and the surplus workers lived happily ever after, assured that Lump of Labor is a fallacy and that there's no such thing as demand satiation."?

But the thing is, it's not labor saving. It's labor displacing. That's not to say we shouldn't try to automate this kind of thing; as you said, it's terrible work. But it's still work that someone depends on to be able to support themselves and their families. And usually, with this kind of thing, there isn't much thought put into how to help those who are displaced by increased automation. Those people still need to be able to support their families.

I'm a big fan of decoupling the historical relationship between having salable labor-power and having an income that can be exchanged for goods and services. If the productive economy can go Full Robot, I think that it should, and governments should start planning how their people can get goods and services as opposed to getting jobs.

That said, I think it's fair to expect displaced workers' countries to take the lead on adaptation strategies for their own voters. Some displaced crop pickers are eligible to vote in the US but I think that the bulk of them are eligible only in countries outside the US. Undocumented immigrants don't bother me but neither do I feel like it's the responsibility of Americans to provide temporary foreign workers, documented or not, with jobs in the US.

Its reports like this that make me believe that Universal Basic Income will be a necessity in the next century as all the low skilled volume jobs are taken by machines

We can regret the loss of jobs to automation, but picking is brutal. Very hard on the body. And it's seasonal, so it won't sustain you all year. I don't think it'll be missed.

I mean, it'll probably be missed by those who rely on it for their livelihood. I doubt they'll be offered much in the way of retraining or scholarship.

And that is the crux of the problem. Advances like robots working alongside humans are very good and healthy to quality of life, but the number of people who are displaced by such advances is unfair, brutal, cruel. The main people who are behind these acts are less considerate.

A fact should be established that these technologies are needed, but the intentions with which they are built is sinister and evil. This leads to robots that are not helping a civilization advance but hurting it.

Optimizations could have been created for the worker, but instead they are created for the manager, for the owner, for the businessman, for the shareholder, and for everyone else who thinks workers are the biggest problem in the business.

Thats what Peter Thiel said about CEOs

The thing is, it's not much of a livelihood or career - you can't do it all year, and it's very much only for the young and fit. And usually imported guest workers.

Citation for "young": https://www.theyworkforyou.com/wrans/?id=2004-06-10.175439.h : average age of guest workers issued permits was 22.

Broader discussion of fruit picking seasonal immigration (which mentions automation too): https://hansard.parliament.uk/commons/2016-11-30/debates/9F8...

The irony is that foraging for food can be lots of fun. The monotony and sheer volume of work a picker does is what makes it such a bad job.

I have friends that have lived, worked and farmed in Watsonville, CA for a long time. My wife's grandfather has also farmed in MI for decades. The consensus is that farm-workers are getting harder and harder to come by. There is a labor shortage. This shortage has encouraged the investments and greater interest in picking technologies.

So to reframe, there are fewer farm workers anyway, due to better border control, lack of interest, or whatever and the picking technologies become more useful.

>>There is a labor shortage. This shortage has encouraged the investments and greater interest in picking technologies.

It's funny how it hasn't encouraged paying people more.

Its tough to raise the price of an agricultural product to accommodate for higher wages. The market sets the price.

Plus its literally backbreaking work. Illegals would rather hang drywall, paint houses or whatever and make 50% more.

And it's an international market. Pickers in America are competing against the cost of labor in developing nations.

There is no shortage of farm workers any more than there is a shortage of gold bars. Employers are simply unwilling to pay the market price.

We read quite a bit on here about robots and machine learning replacing jobs. Yet productivity growth seems to be almost stagnant, last I heard. So judging by the statistics, this is as yet not a real, large scale phenomenon.

It would have been nice of Bloomberg to mention that Brad Goehring is a politician. https://www.facebook.com/pg/Brad-Goehring-118538778165929/ab...

harvest automation is the most difficult and imo, the least urgent. there is a difference between robots assisting human labour and robots replacing human labour.

strawberry picking in japanese hydroponic farms comes closest to decent automation. the robotic gripper can pick the ripest strawberry just in time. this is mostly possible because ripe strawberries can be easily recognised because it completely turns colour. colour sensors are utilised here. i have heard of brix measuring also, but i cant imagine how it can be done.

strawberry picking is brutal in a regular farm and it is hideous for the soil. it has to be picked a tad unripe so it is ripe when it reaches the supermarket or wholesalers. that's why u-pick strawberry farms are so prevalent.

dutch milking robots are the best example of automation in agriculture..altho' its not technically 'agriculture', but rather animal husbandry..but..oh well.

This site has estimates for agricultural workforce over time: https://ourworldindata.org/agricultural-employment/

It is impressive to see how small the fraction has gotten already.

Robots have been replacing farm workers a lot longer than this. 200 years ago, 90% of the population was involved in farming. Today, it's less than 3%.

there is a huge opportunity for robotic in farming. Tech industry completely ignores it because it's not sexy enough!

I'd say it's not a particularly easy issue to tackle, but I can imagine a farmer renting a specialized fleet of robots to pick x crop and then another set of robots for y crop. It will be hard to create a general purpose robot for picking both oranges and almonds, but only a matter of time.

or is it because a fruit picker working for a few $ per hour is just cheaper?

hiring taxi drivers are relatively cheap too, but all big tech companies are investing on self driving cars. and because of all the investments and research we know we will have a working model soon. also investment reduce the cost. I can argue that cheaper and better food is even more important for the society and human race than autonomous cars, but because it is not sexy enough no one really spent money on it!

It may be that innovations that enable shipping of agricultural products from countries with cheap labor will be cheaper than robots.

Is all this investment in self driving cars done to replace taxi drivers? Will food get cheaper if picked by robots? And even if, what the use of it if nobody can pay the cheaper prices anymore, because they don't have work anymore, because their work is done by robots? Do you really think big money does decide by sexiness of the product where to invest? Like some JS Hippster selects a framework?

yes absolutely I believe that is how investments get decided on big corporations. think about it, the only goal of any CEO is to drive the stock price up. investing on sexy projects create hype and drives the stock prices up, the more you get media coverage about your new project and more people here about it, there going to be more investors and stock prices are going higher !

> because of all the investments and research we know we will have a working model soon

This definitely does not follow.

A lot of crops are mechanized. Wheat, corn, many nuts, some kinds of vegetables, some fruits.

But as the article points out there are some crops that are just too delicate to harvest mechanically without destroying too much (or basically all) of it, so it's not economical.

If your choice is $50k for labor on a $100k of crop or a large upfront investment plus $0 labor and only 10% of your $100k crop is harvestable... the choice is obvious.

(all numbers made up)

true. field harvesting machines are fascinating. fruit trees are different. nut tree shakers are already pretty common but they are highly polluting...raising clouds of dust. most people dont know how destructive large scale commercial farming is to the environment. gilroy, salinas, central valley...fresno, modesto, stockton...the air quality is the worst. as is water pollution. and farm workers and their children have the worst diets in the country. and they are worst paid...and most vulnerable. declining quality of life and poor health of an entire underclass of the population IS costly. when the children of farm workers are at a higher risk for developing type 2 diabetes before they hit their teens, its a problem. its not luddite to replace farm workers with machines. i cant even...ugh. we need machines. we need farming robots. it is the right thing to do. i am a tiny tiny farm...but the farms that feed the world sprawl over thousands and thousands of acres. they are ginormous and its a mess, frankly. while many tasks in most farms is mechanised, there is certainly a price for it. the advantage to robot farm workers is that they can put in more hours, can self power themselves(we will see this in the future), clean energy, wear and tear can be repaired.

No. My dad is a peach farmer. Human workers are more expensive than automation, but not only that: they are difficult to find and come with tons of regulation and legal risk.

Peach harvesting machines exist and my dad would love to use them, however they bruise the peaches too much to be used. Current machines simply shake the tree and let all the peaches fall into a net. Also, those machines are single purpose but picking is not the only labor intensive task. Thinning, propping, and pruning also require a lot of labor.

GP has been downvoted, but I think they're right to some extent. There is huge potential for robotics in farming, and although some projects exist I think there is underinvestment relative to other areas. Some of the most advanced and productive farms in the world are less than two hours drive from the heart of Silicon Valley, yet I see no collaboration there.

>Tech industry completely ignores it because it's not sexy enough!

Tech ignores it because the profits are razor thin.

It's hard to improve on sun+dirt+human hands.

I'd pay more for food that has never been exposed to animals.

This includes humans, birds, rodents, bugs... and their manure too of course.

The store counts too. Seal it before it leaves the greenhouse.

> Tech industry completely ignores it because it's not sexy enough!

More like tech journalism ignores it. There's already been a ton of investment in farm automation.

John Deere Autonomous Tractor - https://www.youtube.com/watch?v=xh2O0c9VBO0

Agrobot Strawberry Harvester - https://www.youtube.com/watch?v=RKT351pQHfI

Ferrari RemoWeed - http://www.ferraricostruzioni.com/en/series/remoweed

Agricultural Drones - http://www.precisiondrone.com/

Lely's Automated Milking Machine - https://www.youtube.com/watch?v=jYg0haya28I

The tech industry ignores it because it will never return the home run that VC capital needs.

Merely profitable businesses don't fit into the tech ecosystem very well.

Most of the tech industry is "merely profitable businesses". Unicorns are for VCs and little girls who still believe in stories.

Email puts mail delivery man out of work.

This luddite news clickbait should not have a place on HN.

Showing concern for those who will be without a way to support themselves or their families is not being a luddite.

Historically, the luddite concern was that technology would replace humans with no consideration for what society would do with/for those humans. This was, and assuredly still is, a valid and pressing concern.

Technologies (outside of military) are rarely regarded as intrinsically bad, but very few administrations over the millennia have felt comfortable talking to their citizenry about how they're going to approach massive social and economic changes.

This is in part a failing with governments throughout history, but doubtless also a sad indictment on human nature (difficulty with short vs long term plans, concerns over wealth disparity, preference for inherited social expectations / status quo, etc).

Just gonna put an idea out here: what if food was so abundant it was free? How much cheaper would it be to live? How much could be saved on redistribution schemes like unemployment benefits or Social Security?

50,000 years ago, everybody's job was to find or hunt food. If you didn't your kin-group starved. Then the Ice Ages thawed and we developed agriculture, so only about 70% of the people had to produce food, while the others could specialize in the production of tools or the provision of services like health care. Industrialization and the Green Revolution have made food so plentiful, we can have foods from all over the world available to us at any time of year.

So instead of worrying about the future of yesterday's dead jobs, let's make the cost of living as low as possible.

> Just gonna put an idea out here: what if food was so abundant it was free?

The land it's farmed on has inherent value to someone; there's no amount of labor cost reduction or automation that can bring food costs below the value of the land + whatever raw materials go into production (seed, feed, whatever).

> Industrialization and the Green Revolution have made food so plentiful, we can have foods from all over the world available to us at any time of year.

And yet people starve all over the place; not just in 3rd world countries but here in the USA. So either it's not that plentiful or we have a serious logistical problem.

> And yet people starve all over the place; not just in 3rd world countries but here in the USA. So either it's not that plentiful or we have a serious logistical problem.

It is the latter, in case you were curious. Access to food requires uninterrupted distribution, which is not always the case.

I don't think it's distribution - there is food available in every store I've seen. It's money to buy the food with, or money to buy mobility to get to the food.

They've been playing a PSA lately saying that "40% of the food produced in America goes to waste". So, on average, people in this country are already spending close to double the amount that they should for food.

So, there's distribution from farm to store to home, but there's a non-existent link out of the home into a second-stage distribution pipeline, or something.

This might be the source of that statistic:

https://www.nrdc.org/sites/default/files/wasted-food-IP.pdf

Only about 25% of the losses (ie 10% of the food) are attributed to "consumer losses".

i have a farm and more than 40% never reaches anyone. it is not worth it for me to harvest it even to donate it because the labour would wipe out my profits. but if i let it rot and decompose..make my own compost, there is some value in it for me. isnt that awful? but yea...farming reality. it is what it is..

Is that one of the situations where subsidies make it worth it to grow in the first place? Or where growing it as use for fertilizer makes it worthwhile?

I'd note that the paper that your sibling comment posted breaks down waste into the different segments of the supply chain. The biggest single chunk looks like they're in the "consumer losses" segment, and that the PSAs I've seen are focused on how to interpret "best by" dates, eating leftovers, and buying only what you need right then, rather than trying to blame farms for waste.

Money for mobility is the same as the food being too far away.

How many people in the US starve to death from simple lack of access to food, without other physical or mental problems that prevent them from obtaining food from one of the many available sources?

Almost every major long term famine in the past 100 years was caused by violent conflict or colonial occupation. Those root causes create the logistical problems.

A diet built mostly on rice and beans[0] with the occasional serving of liver, piece of fruit, can of sardines, and multivitamin to make sure you don't end up deficient in something important is already very, very cheap. So cheap that the annual cost of the square footage for the kitchen to cook it in (maintenance, rent/mortgage, tax. et c.) may be a bigger concern.

Granted it's not a very fun diet, but hey, there are people who pay a whole lot of money for a bland food-goo meal substitute (which, admittedly, removes the need for a kitchen).

[0] This prepper site http://geekprepper.org/what-is-the-cost-of-5-gallons-of-bean... (2015) figures over 100 days worth of rice+beans at ~$125, not counting containers to keep it in. Call it $300/year for a ~2500 calorie/day diet (in JUST rice and beans, not counting anything else). That's $700 for other stuff with a budget of just $1,000/yr/person for food.

[EDIT] now that I think about it, you could probably get away without a real kitchen on this diet. Rice cooker or crock pot, microwave, mini fridge, water source. Portable propane burner if you can swing it, for when you have stuff that's not rice+beans. [EDIT EDIT] and you can skip the microwave if you have the burner.

> with the occasional serving of ... multivitamin to make sure you don't end up deficient in something important

multivitamins do have to come from somewhere. And they can be squeezed from actual foods or created synthetically [1].

[1] http://www.precisionnutrition.com/all-about-vitamin-suppleme...

If you got most of your calories directly from corn (instead of indirectly) you could keep yourself going for literally pennies a day. I have a 50lb bag of cornmeal that I got for ~$10 through bakery distributor. About 600g of that would provide sufficient calories for a typical day, so that is a 38 day supply - costing just over a quarter a day.

food is abundant. it can be free. and yet..not. farming isnt 'high profit' by any definition. it is however expensive because you cant throw seeds and eat food. labour is expensive. there is a terrible and tremendous cost to working the soil. and it's invisible too. it is bad for the environment but there is no way to quantify it. everytime you see a tractor rip the soil, that's carbon released into the atomosphere. it is highly dependent on fossil fuels. seeds have become proprietary. you are paying more for distribution and storage..sometimes triple the cost of the food itself. its complicated.

Sooo... the workers still doing the work get paid more than they used to and we create new jobs designing and manufacturing the equipment. HN really needs to starting flagging these poor stupid workers will all be out of work porn.

Diagnosing doctors will be replaced way before migrant farm labor disappears.

Do you really think the proportions of that are going to work out? That laying off a thousand farm workers will somehow generate a thousand replacement engineering jobs (and that these farm workers will somehow get trained to fill them)?

They will not be laid off in such numbers because there is little economic incentive to do so. Machines cost a lot of money. And they require maintenance over time. You have to realize how little these workers make. Paying US rent full time a person could barely afford to feed and shelter themselves on the money these workers make.

We are talking 50 cents a bucket. Pennies per item. There is not a lot of room for automation to undercut.

Workers in decision making and diagnosis type roles cost significantly more and mostly only require software(not hardware) to replace. Think about how much a doctor gets paid per cancer diagnosis? Or per cancer treatment prescribed?

Reply on news.ycombinator.com