Hands-on with Baxter, the factory robot of the future

I recently did some research on robot state-of-the-art, and Baxter is one of the standouts. Baxter combines advanced sensors with an interesting approach to learning a mechanical routine, ostensibly trained by a factory line worker, not a programmer. The founder was also one of the iRobot founders: http://newsoffice.mit.edu/2013/rodney-brooks-rethink-robotic...

This isn't totally uncharted waters. There are similar robots/arms out there, cheaper even than Baxter's ~$20K price tag: http://www.universal-robots.com/

I think the semi-human factory robot is going to be a big robot segment eventually, and part of the path to robots moving around within society. Companies like Rethink Robotics end up working on necessary problems like how not to whack humans with giant robot arms, and make humans feel comfortable with the robots next to them.

My firm does industrial automation consulting. I've seen one of these in the wild at an in-house demo for a furniture manufacturer and we have an auto parts supplier who is looking at using them in production.

They seem to have a ludicrously low weight limit (I remember reading 5 lbs) and don't seem capable of tasks where you need a programmer for automation (welding, quality checking, etc, etc.)

I would not discount the likelihood of managers and owners getting unduly excited and buying ten of them before having to quietly admit to themselves it was a mistake.

I think they may a decent set of sales, but that's more about cutting out the programmer from the process (it's a fairly basic packing robot by the looks of things and if your factory is laid out well enough to fit it in, you will probably have a more specialised robot doing that work anyway. This seems too generic too be effective.

I think the biggest open market here is turning things that are not factories into factory-forms. For example my house is a new build and it was stuffed up (forgot to put in insulation). Now as the CEO of the building company I would rather have put QR codes on all my planks of wood etc, have them photo graphed in place then put the plaster on top, thus being able to record the processes, add analytical to something that is "in the field".

The major benefits will not come from replacing humans with robots but with replacing lack of control with control.

I got a chance to see and play with one of these not too long ago. Cool technology, not really ready for prime time.

Anybody else getting flashbacks to Fallout New Vegas?

This isn't a "factory" robot. Maybe a toy for some hardware startup to waste VC funds on but not something you'd want in a proper factory.