In December 2013,?it was revealed that?Google was?starting a robotics division?headed by Andy Rubin, the founder and former head of Android. Rubin immediately went on a shopping spree, snapping up robotics companies?Schaft Inc,?Industrial Perception,?Redwood Robotics,?Meka Robotics,?Holomni,?Bot & Dolly, and?Boston Dynamics. While Boston Dynamics is the headline grabber?it was basically the US military’s robotics wing?Schaft Inc. is a big deal, too. Its namesake robot recently won the DARPA Robotics Challenge, a disaster response competition created in the wake of the Fukushima power plant failure.
In fact, if you use the DARPA Robotics Challenge as a barometer for the robotics industry, Google’s ownership of the robotics space borders on monopolistic. Through Schaft, Google handily won Track A, the main bring-your-own-robot competition, and Tracks B and C were software-only challenges using the Atlas robot, which is built by the Google-owned Boston Dynamics. There are only a handful of other high-profile players left in the robotics industry: iRobot, which makes some wheeled military robots and consumer robo-vacuums;?Rethink Robotics, a company founded by an iRobot co-founder and the makers of the assembly-line bot Baxter; and the seemingly disinterested Honda Motor Company, the makers of ASIMO.
It’s not entirely clear what Google intends to do with its new robot wing. It’s expected that the company will somehow tackle the business sector first, which is much more likely to drop $10,000+ on expensive robot parts compared to the consumer market. The goal of the division isn’t to be a long-term research project?it hopes to get a product to market sooner rather than later.
One robot initiative that very much?is?a long-term research project is Google’s self-driving car experiment. A self-driving car is really just a hulking, four-wheeled robot that is programmed to follow the rules of the road. Google hasn’t given us a status update on the project lately?the last we officially?heard of the project was in August 2012, when Google said it drove 300,000 miles without the car causing an accident. Legal issues are going to be one of Google’s biggest hurdles, but the company has slowly been making progress: self-driving cars are now legal in Nevada, Florida, California, and Michigan.
In August 2013,?Amir Efrati?a journalist who has a great track record with Google rumors throughThe Wall Street Journal?reported?that Google was exploring commercializing these self-driving cars. Efrati reports that at first the company met with traditional car manufacturers about integrating this technology, but when those meetings didn’t go anywhere, it started talking to major auto-components companies about building its own vehicles. Most auto manufacturers are terrified of self-driving cars for both marketing and litigious reasons, so if Google really wants to create a consumer product, chances are good that the company will need to go it alone.
Besides selling the cars to individuals, Google has also considered the idea of a “Robo-Taxi”?imagine Uber (a Google Ventures investment) without the drivers. It’s illegal to not have a driver behind the wheel of a self-driving car, however, so it’s unclear how Google expects to ever operate a robo-taxi service. Either way, cost is a serious factor. The current autonomous driving gear adds $150,000 to the cost of a vehicle, but Google has been trying to drive the cost down by building its own components.
Google is working on a smartphone with user-removable parts called “Project Ara.” The idea is a totally customizable phone, meaning even the internal parts can be removed by the user.?Each part?like the SoC, camera, screen, and battery?lives in a detachable “module”?and can we swapped out for more or less expensive components. The project was originally started at Motorola, which is slated to be sold to Lenovo. However, Motorola’s R&D department and Project Ara will be?staying at Google. Completely customizing a phone would be interesting,?but having every single part come in its own external case and click into a framework is going to take up a huge amount of space. Creating this without it being laughably huge will be a challenge. We’ll have to see what Google comes up with.