Rethink Robotics' Baxter promises a new era of low-cost computing, one that brings the benefits of robots to small firms and perhaps revitalizes local manufacturing. To prove that this is realistic we have a video of Baxter doing useful things.

Baxter is supposed to be the robot for the rest of us. No longer will the robot be the preserve of the big company in special manufacturing facilities. If you have a small production line then the plan is that you can incorporate a robot and reduce the cost of manufacture. This is one of the great hopes of industrial robotics and it has the potential to re-import jobs that have been outsourced to regions of the world where labor is cheap. The thinking is that if a robot can reduce the unit cost of production by doing the jobs that are appropriate for it, then the humans can be employed at higher rates. The benefits of local production offset the higher labor costs and the robot brings down the overall cost.

Of course all this depends on the robot being able to do the job.

Rethink Robotics has produced some demo videos of Baxter doing simple tasks, and in this video the argument is that the robot is doing realistic factory work:

Yes - Baxter is impressive.

The moment that the pair of arms starts to work independently is slightly spooky - if only I could do that! But...

As you watch the video you can't help but notice that Baxter is slow. Probably too slow for most factory environments. What human watching the conveyor belt sorting doesn't want to rush in and throw the ducks in the water and finish the task so we can move on to something else.

It is true that Baxter can afford to be slower because it can work 24x7 and doesn't take breaks, but my guess is that this isn't enough to give the robot the edge over a low-paid unskilled human motivated to work fast by piecework rates

Perhaps an updated Baxter could achieve the speeds needed to work on the line with humans, but at the moment the best applications for the robot are those where speed isn't an issue.