On a sunny Sunday morning recently, passengers disembarking from a Royal Caribbean ship in South Boston might have thought they’d time-traveled into the future. Set up in a parking lot next to the cruise terminal were enough robots to fill a few sci-fi movies: drones that could take off and land without human involvement, quadrupedal walking bots the size of a golden retriever, and self-driving cars.
The event, organized by the trade group MassRobotics, offered a look at the health of the region’s robotics sector, which has seen some recent successes — one company, 6 River Systems, was acquired for $450 million last month. But there have also been some expensive flops. Three companies founded by high-profile roboticists — Jibo, Rethink Robotics, and CyPhy Works — have all powered down in the past 13 months, taking more than $240 million of investors’ money with them.
Six things are happening in robotics in the Boston area right now.
The warehouse remains hot. The news last week that Amazon Robotics is adding jobs and workspace in Westborough was a ripple from Amazon’s 2012 acquisition of a local robotics startup, Kiva Systems, which now supplies much of the infrastructure that enables Amazon’s massive warehouses to fulfill orders so efficiently. In 2015, a couple of Kiva alumni founded 6 River Systems of Waltham, which also sought to make warehouses more efficient. It was snapped up in September by Shopify, a Canadian company that provides support services to online retailers.
Investors remain cool. Kiva and 6 River may be the two biggest financial successes of the past decade in robotics, says Ajay Agarwal, a venture capitalist at Bain Capital Ventures in Boston who put money into Kiva. “In robotics, there have been a handful of successes and a lot that have not worked,” Agarwal says. Even though his firm did well when Amazon bought Kiva, Agarwal says that he hasn’t “seen a lot that we were excited about” backing since Kiva. Agarwal and Bilal Zuberi, an investor at Lux Capital, agree that many robotics companies are more interested in working on really tough technical problems — often a continuation of their founders’ academic research. Most of the robotics sector’s wins have been “under people who are not trying to commercialize research, but identifying real business opportunities, and only then putting solutions together that address those problems,” Zuberi says. (His firm was an investor in CyPhy Works, a Danvers maker of surveillance drones that ceased operations earlier this year.)
The most consistent local investor in the robotics sector? iRobot Ventures, an arm of Burlington-based iRobot. Since 2014, it has put money into about a dozen startups — including 6 River.
Simplicity sells. Perhaps the hottest-selling bot at the MassRobotics expo is made by a Lincoln startup called MVP Robotics. It’s a 160-pound, $5,500 tackling dummy with a top speed of 16 miles per hour, designed for football practices. There are no artificial intelligence routines running in this robot’s noggin: its actions on the field are controlled by a coach on the sidelines. But it’s a way for teams to practice tackling in a realistic way, without human putting players at risk. And MVP already counts as customers half of the National Football League’s teams, and about 50 colleges. The company’s next product is a mobile dummy dubbed HEKTR that police officers can shoot at with live ammo during training. If your aim is good enough, HEKTR falls over.
Drones for business use may soon take flight. American Robotics is a drone-maker that has built a fully-automated system that’s contained in what looks like a big white crate. It’s basically a drone airport in a box. When it’s time to fly, doors on the top slide open and an aircraft with four rotors rises into the sky. When the drone returns, it can be recharged without human intervention. American Robotics CEO Reese Mozer says the first market they will focus on is agriculture, giving farmers a high-resolution picture of how well their crops are doing on a daily basis. But before American Robotics is cleared for take-off, it needs to prove to the Federal Aviation Administration that its system can detect and avoid other aircraft “without the need for a human pilot or visual observer,” Mozer explains. The company hopes that will happen next year.
Can Boston Dynamics transform YouTube views into revenue? Some of the best-recognized robots in the world right now are produced by Boston Dynamics, a Waltham company owned by the conglomerate SoftBank of Japan.
You know — the dog-like and humanoid bots you’ve seen on YouTube that fall over, get back up, open doors, and generally look like they’re ready to take over the world. Boston Dynamics wasn’t showing off its robots at the MassRobotics open house last month; CEO Marc Raibert says it’s too busy doing testing of a four-legged walking robot called Spot at “customer locations all over the country, including construction sites, an airport, gas and oil facilities, electric power distribution facilities,” and an unnamed entertainment company. (Raibert says there will be a Boston-area demo at the Museum of Science on Dec. 15.)
Zuberi, the venture capitalist at Lux, says he’s still a skeptic when it comes to Boston Dynamics, which spun out from MIT way back in 1992: “I haven’t see any signs that they are building systems that solve any specific problems as yet,” he says.
Boston startups are trying to identify lots more jobs robots can do. One thing that makes robotics such an interesting field to track is the continual search for ways that today’s technologies can perform a useful task.
Root.ai is developing a robot that can tell when fruits like tomatoes or strawberries are ripe — and if so, pluck them without accidentally juicing them. Impossible Innovations is developing a snake-like robot that would be able to inspect the inside of pipes and other tight spaces.
Pickle Robot Co. is building a robot (named Dill, naturally) that will be able to load or unload trucks. And Franklin Robotics, based in Billerica, has already delivered 2,000 solar-powered robots that can whack weeds in a garden. But founder Joe Jones notes that there are 42 million gardeners in the United States, and most of them don’t love dealing with weeds. He wants to make Franklin’s $350 weeding robot “as ubiquitous as the hoe,” he says.
Sounds crazy, until you learn a bit about Jones’ background: while working at iRobot, he led the team that created the Roomba robotic vacuum cleaner. That product is the Big Mac of the consumer robotics biz, with more than 25 million sold.