This is part of the general trend in robots summed up roughly like:
1. Corporations want to save money and are willing to spend up front for it. People will replace personal vehicles far more slowly.
2. A restricted environment is way easier to engineer a robot for vs. arbitrary environments.
3. It is easier to train workers about working around heavy machinery than to teach arbitrary humans to look both ways before crossing the street.
4. It is way easier to get permission on privately owned land than to convince the government for a state or nation wide laws.
I work in autonomous drones and see similar trends in that space too.
Or just buy a self driving vacuum and you have a self driving vehicle in your own home. It isn't nearly as impressive when you remove traffic etc from the equation though.
Once the tractor is out in the field it can be automated rather easily yes. The tech makes sense on big farms with flat fields and no wild animals. Woods and non-flat terrain, different story.
Is there a non-"team green" solution that works well? With all the bad press about Deere these days (including a few HN comments from suppliers to Deere that revealed they deliberately ask for slightly tweaked parts to avoid being compatible with regular parts), I wonder if there's a suitable alternative.
I live in a rural area and some people are selling their used John Deere machines to buy new ones from other brands due to this problem. It won't surprise me if those other brands invest more in that technology in the next years to be on par to John Deere, and it would be really positive if they force a higher level of repairability.
Anyway, those who sell those products probably would also sell the autonomous technology as a service more than a one pay product. If those services are arriving in the car market I doubt that machinery industry would reject earning a periodical income in addition to machine sales.
I actually feel that is a real problem and shows the little things developers don't take into account and why self-driving cars are going to kill a bunch of people until they debug it, with you as the beta-tester.
Self driving tractors and self driving cars have much different safety features. Don't get me wrong, tractors should absolutely be smart enough to handle wildlife like that. But comparing the two is unfair.
What is your opinion on the societal effects of this lack of trust? This default trust on computer systems undermine trust in other humans. How can we hope to live in a community we do not trust? Why would i take into consideration X opinion on foreign trade if this "computer" thing is likely to be less wrong?
An elevator can be considered a purely mechanical, not-social thing.
Traffic can have a social aspect to it, most of us do not really care and consider it too to be a not-social thing.
My point was more about what happens when we take things with social aspects and start to think of them as purely "mechanical" things. At that point having a human do activity "X" will appear dumb, like having a human control an elevator today. Could we still call this future world a society, since there is nothing social about it? Essentially we will dehumanize everything, and to me it seems sad.
I have no issue with driverless cars per se but I am extremely interested in retaining vehicles that are not connected and/or central controlled and are autonomous under my sole control.
Freedom of right to unsupervised movement is as important as free speech in a free society IMO
I'm really surprised autonomous recreational boats are not something yet. There are plenty of canals and rivers with day boats and holiday boats going about slowly around the 3-6mph speed limits which would make ideal candidates as this is a very limited contained environment. At the very least the insurance companies could be insisting on some sort of crash avoidance guidance/control system to avoid boats from getting damaged whilst protecting riverside property.
Isn’t the point of recreational boating that you drive it yourself? Like with turning the meats when barbecuing? We daddies need something important to do. :)
You arent allowed to drink and drive and when people go on a boating holiday that becomes inevitable. Besides it can get quite boring pottering around at 3-6mph on a canal or river, the novelty can wear off quickly especially if its raining and you have to stand at the back steering a canal boat. So why not have some sort of autonomous guidance system? At least it can become more pleasant in bad weather and opens up the holiday season a bit more for those who are happy to sit inside and see the world go by from the comfort of the inside of the boat.
In a way, its why coach/bus holidays exist, the tour guide on the coach points out the sights whilst you relax in the comfort of your own seat, but having your own space, like a boat can also it make it relaxing especially if you dont want to be around other people. A coach holiday can still feel like a train commute into London, but with a seat.
Another area of development which the car manufacturers might want to develop, is radio controlled cars kits with active suspension and aerodynamics with some sort of AI/coding control. Like a Formula1 championship but radio controlled model cars. It could sow those seeds for kids to learn to code and then test their code on a race circuit.
You’re right. At least the option to turn on autopilot would certainly be nice to have. It doesn’t have to be one way or the other.
Re: R/C autonomous toys. I think these already exist. You can build them yourself from a raspberry pi. And I think there are also kits for autonomous robots. Haven’t heard of racing cars, but small robot creatures for indoor use. I don’t remember any names of brands.
Canals (or at least narrow, shallow British-style canals) would be a surprisingly difficult environment to design an automated system for. Vessels whose awkward handling is heavily affected by highly variable wind, water depth and proximity to banks, overhanging branches, floating rubbish and interactions with other boats whose unpredictable behaviour is usually understood by talking to them. Given the delayed response because 20 ton boats have massive inertia, and can be blown sideways even when brought to a halt I'm not even sure reliable crash avoidance is possible. Plus the real hard work (and risk to humans) is the mooring and the locks.
Certainly a semi or mostly automated system would have its advantages, but not necessarily enough to justify the costs of actually training it for a very small niche market dominated by people proud of using obsolete tech.
Do you have any info/links on autonomous shipping ?
Deep sea autonomy should be a .. breeze (ha!). Last mile is obviously a different beast.
Maybe the trend to larger ships makes autonomy just not worth it. Crew cost is fairly marginal. AFAIK shipping is fairly legacy and regulation-bound, with captains filling in excels and emailing them daily to HQ.
EDIT: i recall a Norwegian cargo ferry doing a daily near- shore route. kudos to them
While not canal boats, some yachts do have varying amount of automation. Some have waypoint systems tied into radar/vision collision avoidance systems which will guide the boat through a series of waypoints and attempt to avoid hitting anything on the water. Obviously you're supposed to still be alert and in control of the boat, but automated systems for cruising around a more open area definitely exists.
You definitely wouldn't want to trust it for any place even slightly crowded though. On a big yacht you'll want to respond to changes quite quickly, and many of those automated collision detection systems are far from perfect.
There are many things I'm tired of waiting for in industrial manufacturing and engineering, and driverless vehicles are not really on that list. Even if I think about transportation challenges only, the need for human drivers is not even on my top 10 list of issues.
Heathrow Terminal 5 has autonomous pods to ferry people between the carpark and terminal. It seems you can go a long way by constraining the requirements.
A lot further than self driving cars with PRT in fact.
3D rail routing, factory pallet->shop direct delivery, 300mph point to point access, easy emergency and disabled access, no batteries to manufacture, no real land use issues (just poles in the ground for hanging systems), fast infra build out using prefabricated track, simultaneous electricity, transport, power (solar, on the top of hanging rails) and telecoms build out with the rail infra. Fully decentralised infra once several tracks are interconnected.
Russian company cognitive pilot has been working on the same technology. Their approach was to shift the responsibility regarding some functions of a machine and to increase the performance of harvest and to make the driver's job easier.
Without using AI there have been substantial improvements already - I was home for harvest on my family farm last year, driving a JD S970 (last time I drove was a CTSII about 10 years ago).
40 ft front, gps auto steer, radar controlled front height - cognitively it was 10% of what it was 15 years ago.
My uncle had a new X series with a 60ft front that also slaves the chaser bin for unloading so you’re basically automating the unloading as well - it’s wild.
Our country is all to undulating I think to be even close to being fully autonomous anytime in the next 2.0 years but it will be interesting to see how this all develops.
10 years ago the talk of remote tractors was to have one lead manned machine with 1-3 ‘slaves’ machines following behind so the one operator can manage a much bigger area, it’s interesting that this doesn’t seem to be the way they’re going
Would be interesting if you could make this work similar to SolarCity. If own any plot of land, some company will give you a robot for free that will take care of planting seeds and harvesting the crops, and it'll pay for itself over time by taking some portion of the production.
41 comments
[ 1.5 ms ] story [ 93.6 ms ] threadI work in autonomous drones and see similar trends in that space too.
https://www.automate.org/blogs/the-history-of-robotics-in-th...
Anyway, those who sell those products probably would also sell the autonomous technology as a service more than a one pay product. If those services are arriving in the car market I doubt that machinery industry would reject earning a periodical income in addition to machine sales.
https://www.youtube.com/embed/vU-BuWY7wjw
https://v.redd.it/netuqk4d5qn81/DASH_720.mp4
I actually feel that is a real problem and shows the little things developers don't take into account and why self-driving cars are going to kill a bunch of people until they debug it, with you as the beta-tester.
Freedom of right to unsupervised movement is as important as free speech in a free society IMO
In a way, its why coach/bus holidays exist, the tour guide on the coach points out the sights whilst you relax in the comfort of your own seat, but having your own space, like a boat can also it make it relaxing especially if you dont want to be around other people. A coach holiday can still feel like a train commute into London, but with a seat.
Another area of development which the car manufacturers might want to develop, is radio controlled cars kits with active suspension and aerodynamics with some sort of AI/coding control. Like a Formula1 championship but radio controlled model cars. It could sow those seeds for kids to learn to code and then test their code on a race circuit.
Re: R/C autonomous toys. I think these already exist. You can build them yourself from a raspberry pi. And I think there are also kits for autonomous robots. Haven’t heard of racing cars, but small robot creatures for indoor use. I don’t remember any names of brands.
Not everyone has the luxury of a bow thruster in tight spaces!
Certainly a semi or mostly automated system would have its advantages, but not necessarily enough to justify the costs of actually training it for a very small niche market dominated by people proud of using obsolete tech.
Autonomous shipping on the other hand...
Deep sea autonomy should be a .. breeze (ha!). Last mile is obviously a different beast.
Maybe the trend to larger ships makes autonomy just not worth it. Crew cost is fairly marginal. AFAIK shipping is fairly legacy and regulation-bound, with captains filling in excels and emailing them daily to HQ.
EDIT: i recall a Norwegian cargo ferry doing a daily near- shore route. kudos to them
You definitely wouldn't want to trust it for any place even slightly crowded though. On a big yacht you'll want to respond to changes quite quickly, and many of those automated collision detection systems are far from perfect.
I think/hope thefuture of agriculture is in diverse cropping & mechanical weed removal by light autonomous machines
We already have driverless vehicles, it's just that cars are downright bad, more or less.
Actually, no.
There are many things I'm tired of waiting for in industrial manufacturing and engineering, and driverless vehicles are not really on that list. Even if I think about transportation challenges only, the need for human drivers is not even on my top 10 list of issues.
https://londonist.com/2014/09/a-ride-on-heathrows-self-drivi...
3D rail routing, factory pallet->shop direct delivery, 300mph point to point access, easy emergency and disabled access, no batteries to manufacture, no real land use issues (just poles in the ground for hanging systems), fast infra build out using prefabricated track, simultaneous electricity, transport, power (solar, on the top of hanging rails) and telecoms build out with the rail infra. Fully decentralised infra once several tracks are interconnected.
https://habr.com/ru/company/cognitivepilot/blog/514678/ https://habr.com/ru/company/cognitivepilot/blog/515432/
My uncle had a new X series with a 60ft front that also slaves the chaser bin for unloading so you’re basically automating the unloading as well - it’s wild.
Our country is all to undulating I think to be even close to being fully autonomous anytime in the next 2.0 years but it will be interesting to see how this all develops.
10 years ago the talk of remote tractors was to have one lead manned machine with 1-3 ‘slaves’ machines following behind so the one operator can manage a much bigger area, it’s interesting that this doesn’t seem to be the way they’re going