I recently had the good fortune of attending a talk given by MIT Professor Emeritus Rodney Brooks, who - after a distinguished career as both an academic and an entrepreneur - has now become an entrepreneur full time in order to create and commercialize intelligent robots that adapt to their environments.
Brooks is best known by the general public (at least the part of the general public interested in engineering innovations) for being the co-founder and Chief Technology Officer of iRobot, to which we owe the home vacuum cleaner Roomba as well as autonomous robots used by the military, such as the SUGV, and others. During his talk he regaled us with "tales from the trenches", for instance:
- His team determined that a potential customer would be able to purchase the Roomba for at most $200 (at the time) without double-checking with her spouse first, and the whole design of the Roomba followed from that price constraint. Brooks said that a European competitor, which had designed a robot vacuum cleaner first, has yet to enter the American market due to the significant difference in price between both products, although the latest, most sophisticated version of the Roomba now sells for $700 (according to the iRobot website).
- Brooks also mentioned that iRobot had a product superior to its competitor's for use by the military in one of its campaigns abroad (I don't remember if it was Iraq or Afghanistan), but the company realized that the end users - young soldiers - far preferred the other product because of its simpler interface, and as a result they were not taking full advantage of the iRobot product's capabilities. The company therefore redesigned the product interface to make it more similar to that of a video game, and the soldiers were then able to use many more of the product's features.
Brooks is now the chairman and Chief Technology Officer of Rethink Robotics (formerly Heartland Robotics) where his team has created the robot Baxter, which has the potential of profoundly changing manufacturing. This is because the robots currently in use cannot work side by side with humans and are very expensive and inflexible. Programming such robots takes a long time (several weeks, I think, because it needs to be done by the manufacturer) and thus robots currently in use have not been relevant to nimble companies producing small batch orders.
In contrast, Baxter has been built in such a way that it can adapt to its environment and does not need extensive prior training in order to be operated safely. The Rethink Robotics website touts Baxter as "America's first adaptive manufacturing robot", and you can find a detailed list of "why Baxter is different" here.
This new generation of robots has attracted growing attention in the press, both in print and in blog posts. For instance, the New Yorker has a blog post up on "Why making robots is so darn hard". Baxter itself is the focus of an excellent and thorough IEEE Spectrum article: "How Rethink Robotics built its new Baxter robot worker" (subtitle: "Rodney Brooks's new start-up wants to spark a factory revolution with a low-cost, user-friendly robot".)
The company has so far attracted $62 million in funding and Baxter is now about to ship. As the IEEE Spectrum article points out, "Baxter may not be superfast, superstrong, or superprecise like other industrial robots, but it’s smarter. It uses vision to locate and grasp objects, and you can program the robot to perform a new task simply by holding its arms and moving them to the desired position. Baxter will even nod its head to let you know it has understood you."
The article also does a very good job of explaining Baxter's key selling points: safety of use in an environment with human workers nearby (something that I was surprised to learn is not possible with present-day robots), ease of use and cost: "just $22,000", as opposed to hundreds of thousands of dollars.
Brooks is confident that Baxter will sell well after the necessary ramp-up period, in particular because he believes outsourcing manufacturing to the country that happens to be cheapest at the moment is not sustainable, since labor costs in that country inevitably increase with time.
The manufacturing jobs that left the US decades ago are not about to return (or at least they won't be performed by humans), but hopefully keeping manufacturing capabilities in the US will promote different types of manufacturing jobs that will foster continued US competitiveness and innovation.
A quick demo posted on YouTube, although not as fascinating or extensive as the video Brooks showed during the talk I attended, offers a good overview of what makes Baxter so different from earlier generations of machines. Expect to see more of Baxter as Rethink Robotics's customers take delivery of their new robot worker.