Editor’s Note: Christine Magee is an editor for CrunchBase.
If Hollywood’s depiction of artificial intelligence were accurate, we would be falling in love with operating systems, sharing our homes with Stepford wives, and fending off cyborg attacks by now. While movies like Ex Machina and Her stoke the fears and desires of our imaginations, new innovations in robotics and artificial intelligence are bringing our visions of tomorrow closer to today.
You can now take a ride in a Google autonomous car, order dinner from a robot butler at a hotel in Cupertino, and buy a personal quadcopter drone for under a thousand bucks. These robots aren’t quite the cylons from Battlestar Galactica or the space bots from Interstellar, but there’s no question that we’re getting close.
“We’ve seen an order of magnitude of improvement in terms of capability and cost decrease from a robot system in 2005 compared to a robot system in 2015,” says Andra Keay, managing director of Silicon Valley Robotics.
“We saw that same improvement in the decade prior, but the impact was only felt at the very expensive end of town — in defense robotics or hospital surgical suites or dairy farm systems that cost a million dollars,” says Keay.
Now that the contributing technologies are inexpensive and efficient enough, robotic technology is being brought to the consumer market, and venture investors are striving to be at the forefront of the innovation.
According to CrunchBase data, VCs poured $538 million into robotics companies across 104 funding rounds in 2014. They’re on pace to beat that total this year, with $150 million already captured across 22 venture fundings in less than three months.
[protected-iframe id=”9737e95759bfcaa9499dadc6144e2f80-24588526-50862801″ info=”https://public.tableau.com/views/Robotics3_15/ByYear?:embed=y&:showTabs=y&:display_count=yes&:showVizHome=no” width=”680px” height=”500px” style=”border: 0px;” scrolling=”no”]
Investors are especially bullish on consumer-facing robotics innovation. Two of the biggest rounds in 2015 went to 3D Robotics, an open source UAV company, and Jibo, a company building the first interactive robot for the home.
“The rapid decrease in cost of sensors that allow robots to see what’s around them is bringing robots out from behind the big metal cages in heavy machinery areas into environments where they’re really interacting with humans,” says Wheeler.
Google Ventures has backed 6 robotics companies to date, according to CrunchBase, that range from drone operating platform Airware to Transcriptic, a company building robotics for biology labs. One of Wheeler’s recent bets, Savioke, has created a delivery robot that is currently making its debut in a handful of Silicon Valley hotels.
“In a house you might have two or three bedrooms and serve five or ten meals a day, but in a hotel you’ll have 300 bedrooms and serve thousands of meals per day — that repetition is something that lends itself to robotics,” says Steve Cousins, founder and CEO of Savioke.
Prior to founding Savioke, Cousins headed up robotics research lab and incubator Willow Garage where he oversaw the creation of the robot operating system (ROS), the PR2 robot, and helped launch 8 robotics companies — two of which have since been acquired by Google.
“We’ve built almost as simple of a robot as we can get in order to push on the other aspect of robotics, which is moving from demos, where robots typically are, to deployment,” says Cousins.
Drones, or flying robots, have demonstrated just how quickly robotic technology, once deployed, can achieve mass consumer adoption. Two years ago, it wasn’t possible to build a small drone because the computing horsepower was too heavy to lift with propellers. Now consumer-friendly drone maker DJI is on track to become the first billion dollar drone company with over $500 million in earnings last year alone.
“A lot of people are starting to wake up to robotics because they’re seeing what’s happening with drones and autonomous vehicles right now,” says Rob Coneybeer, founder and managing director at Shasta Ventures. Shasta recently participated in a Series A round for stealthy service robotics company Fetch Robotics.
“Multiple enabling technologies are crossing the threshold at the same time, so the advances are coming along far faster than anybody realizes,” Coneybeer says.
[protected-iframe id=”377fd3978a381b9e35b7330794ca28b9-24588526-50862801″ info=”https://public.tableau.com/views/Robotics3_15/Bubbles?:embed=y&:showTabs=y&:display_count=yes&:showVizHome=no” width=”680px” height=”500px” style=”border: 0px;” scrolling=”no”]
It’s not quite on par with the drone frenzy, but consumers have already proven eager to introduce new robotic technology into their homes. Social robotics startup Jibo raised $2.2 million and pre-sold 4,800 family robots in its record-breaking Indiegogo campaign last year, prior to raising $28 million in venture funding this January.
“Jibo is kind of like a cross between a tablet and a puppy — yes there’s robotics involved, but it’s also a companion that sits on the kitchen countertop and interacts with the family,” says Bruce Sachs, partner at CRV and investor in Jibo.
Trusting Jibo to schedule your appointments, teach your child math, and act as your personal fitness coach — and all with a personality that develops as it learns about you — is quite a jump from installing a Nest thermostat or Canary security system on your wall. The amount of confidence consumers have in robots is clearly rising.
But does this mean that robots are going to have the capacity to outsmart us within the next decade, as science fiction would have us believe?
“There is some truth to that, but the problems and the benefits are absolutely overstated and misunderstood. The impact of the technology is vastly overestimated in the short term and grossly underestimated in the long term,” says Keay.
That’s because we tend to associate artificial intelligence with the full realm of human capability, when robots working in the factory or even in the home require a very small amount of AI in comparison. Machines excel in structured environments where they can perform repetitive operations, but performance in unstructured environments is much more challenging to engineer. The latest breakthroughs in computer vision, for instance, allow computers to identify two people having a conversation in a picture — something a human toddler can do easily.
“There’s a gap on the consumer side of robotics because you have this expectation that comes from science fiction movies that these things can do absolutely everything, but in reality the most popular consumer robot is the Roomba,” says Wheeler.
Yes, the vacuum cleaner. At the rate robotics innovation is going, chances are this won’t be the case for much longer.