Slashdot Mirror
RightHand Robotics Automates a New Type of Warehouse Work: Recognizing, Picking Up Items From Boxes (qz.com)
An anonymous reader quotes a report from Quartz: A startup called RightHand robotics recently began piloting technology that automates a task robots have previously struggled to master: recognizing and picking up items from boxes. RightHand can't say which companies are part of its pilot project and Amazon didn't reply to a request for comment. But the new technology could help the ecommerce giant with a problem that has long vexed it. Like robots elsewhere, Amazon's robots retrieve entire shelves and transport them to humans who pick out items from them. They can find and move a shelf that holds a box of shirts, but they aren't capable of removing the single shirt from that box to be packed into an order. In order to pick items from boxes, robots need to master the more complex task of identifying a wide range of objects and adjusting their grips accordingly. RightHand robotics, which was started by a team of researchers from Harvard Biorobotics Lab, the Yale Grab Lab, and MIT, built a solution called RightPick that, according to co-founder Leif Jentoft, can pick items at a rate of 500 to 600 per hour -- a speed on par with a human worker. It uses a machine learning background and a sensitized robot hand to recognize and handle thousands of items.
Smash every robot until we get our Basic Income.
Such as these gentlemen
I can't be the only one who thought of wanking...
Losers! Didn't keep their skills up. Good thing the future is bright for devops and always will be. There's no way automation can combine developing and operating and marketing and public relating and accounting and human resourcing all into one role and fire all the devops. Never!
My understanding is that the main job for humans in Amazon warehouses is for 'pickers', that these machines are claimed to be able to replace (no word on accuracy, however). The article mentions that packing items in boxes is still done by hand, and I imagine loading/unloading trucks is still done with humans. However I can foresee completely-automated Amazon warehouses in the near future. With self-driving trucks, and completely-automated factories, there will likely soon be some products whose packaging are unseen by a human until they reach a consumer's doorstep. Completely-automated retail. From what I could find here, at least 10.1 million Americans work jobs that'd be replaced with automated retail. The American Trucking Association claims 3.4 million American truck drivers. So that adds up to 13.5 million jobs between retail and trucking, add in other driving jobs and it'd be 10% of all jobs.
For comparison, about 2.5 million new jobs are created in the US each year. In the unlikely event every driver and retail supply chain worker were laid off at once, it'd take ~5 years for new jobs to be created to absorb them (assuming an equal number of vacancies.) That's ignoring the fact that many of these 'new jobs' are in the driving and retail sectors. Another 5.4 million Americans work as food preparers/waiters; as minimum wage increases I wonder how many restaurants will increase automation. I know many restaurants won't fully-automate due to tradition or being high-class, but most restaurants aren't too high-class and will automate if it's either that or go out of business.
Corruption is convincing someone that the selfless ideal is the same as their selfish ideal.
At least, having an AI solve real world problems with advanced sensory feedback and complex appendages is doing more for robotics than silly animatronic detours like https://www.youtube.com/watch?....
Imho, these kind of solutions is what is the real driving force of the future of robotics. Now to have a coffee and think about how our warehouse sized, multiped overlords will shape the future.
leftie persecution. they're going smash lefties till they use their right hands.
there goes my best job prospect for when i lose my bus driving gig.
Robots that can see. And respond in sensible ways.
"Bin Picking" requires recognizing which objects are which, and what their orientation ("pose") is. Then plan a way to move to collect them.
That is an order of magnitude more sophisticate than simply moving in rigid, predefined ways to work on things that have been precisely positioned in advanced.
It opens up whole new fields of automation.
And it is not that new, bin picking robots have been around for a while. They are getting better. This is only a story because Amazon is doing it now.
There is a nerd that decided he wanted to enter the exciting future of robotics, yet he would much rather prefer becoming a graduate of the Yale Grab Lab so he can, you know, pick up chicks instead.
WARNING: Smartphones have side effects--most of them undocumented.
It is always amazing how people get so worked up over this. The fact is, the current state of robotics is a joke. Yes, Amazon robots can locate a shelf of products and retrieve it (as long as the shelf is uniform, and labeled clearly). That is releatively easy to do. But pick an arbitrary item out of a box and then pack it? Very very hard. In fact, might not be possible to get it to work reliably. CPU technology isn't going to progress in leaps and bounds anymore, so we might be getting close to the fastest digital processors we will ever see. Of course people scoff and say "Moore's Law", but the fact is that Moore's Law is dead. It is obvious by just looking at the current state of computing that it is dead. That is why there is a rush to multi-core. Everything is dependent on progress of digital processors. And they haven't been progressing at the same rate as they have been historically. Not even close.
A simple digital scale can sort many items by weight. An XL shirt for example will have less weight than a size M shirt. So if the shirts are stored such that all are the same except by sizes the robot can work its way through the bin until the correct size if located. There have been several other useful ways to sort items for quite some time. But the real question is whether robots can sort items that are very diverse and mixed together. So far that is difficult. But go to a junk yard that has one of those machines that can grind a car into small chunks and watch them pick up the steel pieces with an electro magnet and leave all else behind. It is efficient but not perfect as a tiny bit of cloth could be impaled upon a sharp piece of steel. Sorting can be easy or difficult depending upon each situation.
or till you go to jail / prison and you get better doctors then the ER
"But the new technology could help the ecommerce giant with a problem that has long vexed it."
And that problem is....human employees.
Human employees are always demanding stuff like food, shelter, bathroom breaks, medical care, adequate lighting, temperatures above freezing....if we could just get rid of them everything would be wonderful and we'd be living in Utopia!
Just cruising through this digital world at 33 1/3 rpm...
...why things from the vending machine cost more?
"There won't be any jobs for machinists or assemblers. There won't be jobs for anyone. There won't be anyone working except billionaire Bezos telling you to buy shit from robots. When you can't afford to buy, the economy won't need you anymore ..."
I made this parable in 2010 that is about a world that develops along those lines:
"The Richest Man in the World: A parable about structural unemployment and a basic income "
https://www.youtube.com/watch?...
A 21st century issue: the irony of technologies of abundance in the hands of those still thinking in terms of scarcity.