The smartphone revolution that started in earnest with the launch of iPhone (s aapl) five years ago is having unintended consequences on the future of technology, and perhaps the economy. The latest example is today’s announcement of Baxter, a new manufacturing robot from Rethink Robotics, a Boston-based startup co-founded by robotics pioneer Rodney Brooks. And it was one of the reasons I got on the phone with Brooks, who is sometimes called the “bad boy of robotics.”
Our conversation, though limited to half-an-hour, spanned many topics but mostly centered around the role robotics can play in bringing back manufacturing to America. Brooks was on the faculty of MIT as the Panasonic Professor of Robotics and is also the man who started iRobot (s IRBT), the company well known for its household chores robot, Roomba. It was during the process of building iRobot when he realized that robots were going to be key for rethinking and re-imagining low-cost manufacturing in the twenty-first century.
During his Roomba-phase, it became quite clear to Brooks that low-cost manufacturing economies often run into the ceiling of rising standards of living. From Japan to South Korea, from Taiwan to China and more recently to Vietnam — the low cost manufacturing hot-spots continue to shift and there is nothing wrong with the shift, Brooks argues.
However, what happens when the world runs out of shifting manufacturing bases such as those in Asia? Another thing on the mind of Brooks: oil price spikes were adding a fuel surcharge on the low cost products. And there is always the looming threat of intellectual property being compromised. Brooks points out that there are a lot of Roomba replicas. “The big question to me was, ‘can we make low cost goods in the US?'” he said. “It is clear we can’t pay the same wages as places like China.”
It was this backdrop which lead him to start Rethink Robots in 2008 to work on manufacturing robots. It was originally called Heartland Robotics and has raised $57 $62 million in funding from the likes of Jeff Bezos’ Bezos Expeditions, Highland Capital Partners, Charles River Ventures, Sigma Partners and Draper Fisher Jurvetson.
Cheap but not cheesy
Today when we think of manufacturing robots, we quickly recall giant robots that are installed in the car plants inside body shops and paint shops. They cost millions of dollars and are programmed by specialists to do very specific tasks. They are like the mainframes of manufacturing robots.
And that is why the development and release of Baxter is so interesting. It is an inexpensive (sub-$25,000) semi-anthropomorphic line robot that is relatively simple to program. It is very flexible. And it has the potential of reshaping the manufacturing processes — first in mind-numbingly boring tasks such as packaging and removing things off a conveyer belt. “It doesn’t have the dexterity to build an iPhone, but it sure can package an iPhone,” Brooks quipped.
As technology has become more pervasive in our lives, most of us (and by that I don’t mean the early adopters) have gotten used to working with technology without needing manuals. The popularity of touch-based smartphone and society’s growing ease of using search and Facebook means that today, if you make robots simple enough, it is less likely that folks are going to be overwhelmed by them. The trick is to make them augment the humans in the manufacturing process.
What about society’s natural fear of robots and them taking over the world? Brooks laughed, and then pointed out that just as electric drills have made the jobs of construction workers easier, the same goes for these manufacturing robots. Just as computers opened up new opportunities, Baxter and its descendants are going to create new opportunities. Brooks knows it is going to take a long time.
“Simple tasks today,” Brooks said. More complex task will follow in time, thanks to the marriage of hardware and software, he added. In thirty years we have gone from mainframes to computers in our pocket and soon they will be embedded into our bodies, Brooks says. Robots are going to go through the same curve as well.
Smartphones say hello to robots
Like Brooks, I am a firm believer in a future where such machines start to assist in our daily tasks, though the final shape and form of these machines might look entirely different. Will robotics progress fast enough to bring manufacturing back to U.S.? Who knows. After all, we are still not clear as to what kind of manufacturing base we want to build. But the future is exciting and full of possibilities nonetheless.
What is even more exciting — well, at least to me — is that the road to this robotic future is littered with billions of smartphones. The reason why we can build robots like Baxter today is because of the falling prices of sensors and other components. Before the iPhone rolled around, phones didn’t use that many chips. Apple came along and made it normal to demand gyrometers pyrometers, accelerometers, digital cameras, touch and other such sensors.
The growing number of smartphones — a billion shipped by 2016 — has helped the cost of making these sensors and mobile processors decline at a dramatic rate. The chips insider are getting beefier and more capable. It is Moore’s law at work, only at gigantic scale, as my colleague Kevin Tofel wrote last year. And that the reason why Baxter is so cheap to build, because “robotics doesn’t have to work hard to get scale,” he points. Baxter uses nine ARM-cortex chips that are made by Texas Instruments (s TXN) to work and they are all getting cheaper because of the cellphones.
Amazing, isn’t it? And here we were thinking that iPhone and its Android brethren were no more than tools for sharing photos and sending inane status updates.
What would it be called? An iRobot ??
Lol. That name is taken.
Since when has the iPhone had a pyrometer?
I left my iPhone on top of a toaster oven briefly before realizing my mistake. When I picked it up the home screen was displaying a caution message that it was overheating and needed to cool off before being used. They don’t seem to advertise it, but it has one. It may be similar to the moisture sensor(s) that detect if an iPhone gets soaked. The purpose may be more to help for Apple to avoid repairing or replacing water and heat damage on warrantied products than for the user, so there’s little incentive for Apple to advertise the presence of those sensors.
Oops. Sorry about that. I guess sometimes autocorrect does that 😉
I’ve been telling anyone who cares to listen for the past couple of years that robotics is Apple’s next wild frontier. They’re acquiring all the skills necessary to eventually mass produce affordable, reliable and most importantly, desirable, robotic devices for the consumer market. I have no idea what their use would be but then Apple is the master of producing goods which the rest of the market did not see a need and then dominating it.
Of course this is a long term prediction, I’m guessing no less than ten years from now – surely no-one expects them to still just be making computers and smartphones in 2022?
I respectfully disagree with the premise herein. The most expensive parts of a robot are the motors, bearings and other mechanical parts. These parts have become cheaper with time but not because of smartphones. Processors are rarely amongst the high cost components (15 years ago, I worked for a company that made industrial robots, the microprocessor cost $45 while the robot sold for $35k). Furthermore, the sensors used in a smartphone are rarely essential sensors in a robot. Instead, encoders are. They are not ultra expensive but the cost increases with degrees of freedom. These sensors are also not found on smartphones.
He did say it was sub-25,000 so it’s not like he was saying they were pocket change. Moreover, the industrial robots of 15 years ago are not the type of robots this article is talking about. The author took a whole paragraph discussing why these robots are different from robots of the past. The point is that while they are simpler, because of the the smartphone revolution they can be programmed to do more stuff because there’s a proliferation of not just cheap but repurposed sensors that can be used in new ways that were never even considered 15 years ago.