Willow Garage Blog

This is a weak article. First, it ignores the existence of commodity robots. Something like 4 million Roombas have been sold. They are easy to use commodity robots that sell in volume from Best Buy.

Second, many people, going back to the early 20th century, have imagined buying cheap general purpose robots, just as they have imagined buying jet packs and holidaying on space stations. Just because I imagine it doesn't make it feasible. The reason there aren't many autonomous-robot startups is that no one knows how to do the AI. It may make sense to you that you ought to be able to put together robots like you put together a spreadsheet, but that's not how it turned out to be. The article admits this, describing day to day tasks as "incredibly complex". Yet then asks us to imagine the cheap, capable robot.

Third, the "mysterious" capital behind Willow Garage is Scott Hassan, an early and thus very rich Googler. It says so on the WG web page [http://www.willowgarage.com/pages/about-us/history].

Fourth, Microsoft has Robotics Studio, a package that competes with ROS in many respects. It is exactly a "robot operating system" on top of Windows. Researchers pretty much ignore Robot Studio, but why does this journalist do so when he says that "what we need Microsoft [...] to do is build an operating system....". It has been trying for years. Thanks for the advice, though.

A low-quality piece for Spectrum, picked up from a blog.

I don't see how this will work in 3 years time. Cheap robots today (<100K) suck at manual dexterity tasks. It took quite a lot of work to get a PR2 to fold laundry (and some of the best robotics people in the world worked on this). The kinds of tasks I imagined Foxconn workers performed all require significant manual dexterity.

For an amazing use of robots in a factory environment, I would suggest a tour of BMW's Munich factory. They have a whole bunch of specialized robotic arms that do things like welding, lifting, painting, etc. Their factory still employs humans (albeit fewer humans). In BMW's case, their use of robotics makes sense. Foxconn, not so much.

Finally, there are some completely automated car factories. I thought Hyundai's US factory manufactured cars solely with robots. However, I believe they have a higher defect rate (partially due to this process).

Edit: Link to the PR2 laundry folding robot - http://www.willowgarage.com/blog/2010/04/02/towels-uc-berkeley

I think Hackers will have the last laugh after all. When hackers finally build an AI, it will also be the first soul. An intelligent mind decoupled from a body and thus capable of being immortal. The first step in this direction was the stored program concept itself which was best articulated by Von Neumann (http://qss.stanford.edu/~godfrey/vonNeumann/vnedvac.pdf) and realized in the EDVAC. This allowed programs to be decoupled from hardware. Dennis Ritchie, whose loss we all mourn, made what I consider the second big step. Decoupling large software systems from hardware, via a portable operating system. It will take a while (10-20 years?), but maybe something like ROS (http://www.willowgarage.com/pages/software/ros-platform) will someday become a full mind that is still decoupled from any particular hardware...a soul!!!

Sure. There are quite a few. Here is a non-comprehensive list:

http://www.bnrobotics.com/
http://www.resquared.com/
http://www.aethon.com/
http://www.redzone.com/
http://www.seegrid.com/
http://www.bostondynamics.com/
http://www.irobot.com/
http://www.cyphyworks.com/ <-- started by some of the original iRobot folks
http://www.mobilerobots.com/
http://www.willowgarage.com/
http://www.evolution.com/
http://www.sensiblemachines.com/
http://astrobotictechnology.com/
... etc ...

NOTE: Some of the companies are further along than others but all of them (to the best of my knowledge) started as startups and/or university spin-outs (which is a subset of startups). You can normally find startup robotics companies near/around the schools with the best robotics programs (i.e. CMU, UPenn, Stanford, MIT, etc.)

Today Willow Garage announced that eleven (rather than the original ten anticipated) PR2 Beta robots, with a total value of over $4.4M, will be loaned out to academic and research institutions worldwide to develop a slew of impressive capabilities over the next two years. The recipients include 7 US-based institutions, 3 European, and 1 Asian. The final list is a panoply of robotics specialists: University of Freiburg (Germany), Bosch, Georgia Tech, KU Leuven (Belgium), MIT, Stanford, TU Munich (Germany), UC Berkeley, U Penn, USC, and University of Tokyo (Japan) -- full details can be found in the Willow Garage press release:

http://www.willowgarage.com/blog/2010/05/04/pr2-beta-program-recipients

Some smart robotics startups:

MekaRobotics (by Aaron Edsinger, MIT star behind Domo)
http://mekabot.com/
http://people.csail.mit.edu/edsinger/domo.htm

Willow Garage (former early Google employee founder)
http://www.willowgarage.com/

AnyBots (Trevor of YC's company)
http://anybots.com/

Q Robotics (early stars from iRobot that liked startups so much, they needed to leave and start another once iRobot went public)
http://qrobotics.com/

You don't really need to know anything about robots to work on them. The best groups are made up of specialists that understand a bit of everything, but are essentially software|hardware|electrical|systems engineers at their core.