4 ms·

The automation is here and is used everywhere right now if it makes economically sense. Good overview of the pricing here: https://blog.robotiq.com/what-is-the-

by lnsru 7y ago

The automation is here and is used everywhere right now if it makes economically sense. Good overview of the pricing here: https://blog.robotiq.com/what-is-the-price-of-collaborative-robots https://blog.robotiq.com/what-is-the-price-of-collaborative-... Basic robot $50k, advanced application - $100k. One can hire lots of workers for the money. Especially during recession.

devit 7y ago
Aren't these robots so expensive only because they are sold in low volumes? It seems to me that if mass produced, they should at lesat cost less than entry-level cars, since they are a similar size/weight and simpler, so less than 15k. And it also feels like they should cost a similar amount to premium versions of appliances like dishwashers/washing machines/etc., so in the 1-5k range.
- the8472 7y ago
  Robots are mass-produced. Fanuc press material says they're doing 7000 robots arms per month (that's two of their locations, they have more doing other industrial automation).
m_ke 7y ago
People are working on cheaper arms and we should soon have ones for under $10K. Ex: Pieter Abbeel from UC Berkeley https://news.berkeley.edu/2019/04/09/meet-blue-the-low-cost-human-friendly-robot-designed-for-ai/ https://news.berkeley.edu/2019/04/09/meet-blue-the-low-cost-... https://github.com/berkeleyopenarms https://github.com/berkeleyopenarms
- ipsum2 7y ago
  I don't believe the github has the STL files or instructions on how to actually make the robot, just the API and simulations are open source.
  - m_ke 7y ago
    Abbeel's new robotics company might be working on commercializing that prototype (https://covariant.ai/ https://covariant.ai/). I remember him mentioning the project in his robotics class, found it: https://youtu.be/xWPViQ6LI-Q?t=4075 https://youtu.be/xWPViQ6LI-Q?t=4075 he claims it will be around $5,000
    - ipsum2 7y ago
      That's disappointing, it would be nice to see a true open source robot arm. I believe Covariant is separate from the Blue robot arm though.
      - botwriter 7y ago
        https://www.youtube.com/watch?v=cfJn7T4D-6k https://www.youtube.com/watch?v=cfJn7T4D-6k Maybe of interest
- lnsru 7y ago
  This arm still needs accessories, installation and maintenance. Maybe it will end up at $30k.
  - redbeard0x0a 7y ago
    $130k if its for "medical use"
lostapathy 7y ago
"One can hire lots of workers for the money. Especially during recession." When you hire a person, you have to pay them continuously for their work. Once you pay for a robot, you don't owe it a salary. There's obviously other factors - ongoing costs to keep robots going, but also robots never call in sick and can sometimes work faster or more precisely than a numan.
- twic 7y ago
  You have to depreciate that purchase. How long do these robots last? And, as a sibling comment points out, you have to pay for maintenance and operating costs. The capital price of robots is kind of irrelevant, what's the annualised TCO?
  - cookiecaper 7y ago
    And automated systems require babysitters. How long does a 100% robot-controlled factory operate when its supervisors are in quarantine? While it's true that the robots aren't concerned with infection, I doubt that a pandemic-style shutdown can really be fully averted, even if the manufacturing process is "100%" automatic.
    - tialaramex 7y ago
      One human maintaining a factory full of robots is obeying social distancing constraints. They can't infect the robots, and the robots can't infect them. Unlike a factory full of low-paid workers this is actually very low risk for the spread of infection. We actually get to see a small model of this on farms. Once upon a time every farm had dozens or even hundreds of workers. Whether planting and harvesting corn or milking cows, it was all done by hand, and it took a lot of people working in close proximity. Today's farmer has a lot of automation, which means far fewer people. One person in a harvester can harvest a field of wheat, maybe a truck drives out to them once in a while, exchanges hand signals, nobody's close enough to get infected. A couple of people can handle milking a herd of cows, they touch the cows but no need to get close to each other. No problem keeping appropriate distances. Not every farm is like that, it's going to be a problem for fruit harvesting for example where you'd usually bring in a lot of temporary workers once a year. But these systems don't have to change everything. If there are six factories in your city, and three of them remain unchanged but the other three just become a bunch of robots, with the only human on site one mechanic and a security guard that's an enormous change.

by lnsru 7y ago

devit 7y ago

Aren't these robots so expensive only because they are sold in low volumes? It seems to me that if mass produced, they should at lesat cost less than entry-level cars, since they are a similar size/weight and simpler, so less than 15k. And it also feels like they should cost a similar amount to premium versions of appliances like dishwashers/washing machines/etc., so in the 1-5k range.

the8472 7y ago
Robots are mass-produced. Fanuc press material says they're doing 7000 robots arms per month (that's two of their locations, they have more doing other industrial automation).

m_ke 7y ago

People are working on cheaper arms and we should soon have ones for under $10K. Ex: Pieter Abbeel from UC Berkeley https://news.berkeley.edu/2019/04/09/meet-blue-the-low-cost-human-friendly-robot-designed-for-ai/ https://news.berkeley.edu/2019/04/09/meet-blue-the-low-cost-... https://github.com/berkeleyopenarms https://github.com/berkeleyopenarms

ipsum2 7y ago
I don't believe the github has the STL files or instructions on how to actually make the robot, just the API and simulations are open source.
- m_ke 7y ago
  Abbeel's new robotics company might be working on commercializing that prototype (https://covariant.ai/ https://covariant.ai/). I remember him mentioning the project in his robotics class, found it: https://youtu.be/xWPViQ6LI-Q?t=4075 https://youtu.be/xWPViQ6LI-Q?t=4075 he claims it will be around $5,000
  - ipsum2 7y ago
    That's disappointing, it would be nice to see a true open source robot arm. I believe Covariant is separate from the Blue robot arm though.
    - botwriter 7y ago
      https://www.youtube.com/watch?v=cfJn7T4D-6k https://www.youtube.com/watch?v=cfJn7T4D-6k Maybe of interest
lnsru 7y ago
This arm still needs accessories, installation and maintenance. Maybe it will end up at $30k.
- redbeard0x0a 7y ago
  $130k if its for "medical use"

lostapathy 7y ago

"One can hire lots of workers for the money. Especially during recession." When you hire a person, you have to pay them continuously for their work. Once you pay for a robot, you don't owe it a salary. There's obviously other factors - ongoing costs to keep robots going, but also robots never call in sick and can sometimes work faster or more precisely than a numan.

twic 7y ago
You have to depreciate that purchase. How long do these robots last? And, as a sibling comment points out, you have to pay for maintenance and operating costs. The capital price of robots is kind of irrelevant, what's the annualised TCO?
- cookiecaper 7y ago
  And automated systems require babysitters. How long does a 100% robot-controlled factory operate when its supervisors are in quarantine? While it's true that the robots aren't concerned with infection, I doubt that a pandemic-style shutdown can really be fully averted, even if the manufacturing process is "100%" automatic.
  - tialaramex 7y ago
    One human maintaining a factory full of robots is obeying social distancing constraints. They can't infect the robots, and the robots can't infect them. Unlike a factory full of low-paid workers this is actually very low risk for the spread of infection. We actually get to see a small model of this on farms. Once upon a time every farm had dozens or even hundreds of workers. Whether planting and harvesting corn or milking cows, it was all done by hand, and it took a lot of people working in close proximity. Today's farmer has a lot of automation, which means far fewer people. One person in a harvester can harvest a field of wheat, maybe a truck drives out to them once in a while, exchanges hand signals, nobody's close enough to get infected. A couple of people can handle milking a herd of cows, they touch the cows but no need to get close to each other. No problem keeping appropriate distances. Not every farm is like that, it's going to be a problem for fruit harvesting for example where you'd usually bring in a lot of temporary workers once a year. But these systems don't have to change everything. If there are six factories in your city, and three of them remain unchanged but the other three just become a bunch of robots, with the only human on site one mechanic and a security guard that's an enormous change.