The road to home robot is full of dangers. The number of success stories that have been delivered can be counted on one hand. The reasons for this mass disconnection are nuanced and complex – like the inside of our homes. Twenty years after the arrival of the first Roomba, the robot vacuum cleaner has come to seem like a fluke – more the exception than the rule.
Aaron Edsinger, the former Google Robotics executive who now serves as CEO of Hello Robot, isn’t trying to build the universal home robot — at least not now. The Stretch Robot series (not to be confused with Boston Dynamics’ truck unpacking robot of the same name) is a platform the company hopes to build around the next generation of home robots. Watching it cruise around a house in video demos brings to mind Nvidia’s line of reference robots.
The recently announced Stretch 3 is a robot with a wheeled base and height-adjustable handle. In the promo video, you’ll see a couple of Stretches cruising around a house, making beds and unloading the dishwasher — just like people have long dreamed of in a home robot.
There are, however, two very important caveats. First is the $24,950 price tag. As someone who’s been known to complain about high-end Roombas that topped $1,000, it’s hard to imagine paying the cost of a low-end new car — especially given the system’s shortcomings for consumers.
This brings us to point two: The system is tele-controlled. There’s nothing wrong with teleop per se, of course. I’ve said it many times. But the one-to-one human-to-robot control scenario isn’t viable — especially at home, which you probably don’t want to open up to whoever ends up on the other side of the camera.
One place where teleop is great is in the robotic learning process. This is where reinforcement learning comes in — walking the robot through the process of performing tasks in different scenarios. That’s what Tesla is probably doing in that recent video of Optimus folding clothes — even if the company didn’t initially seem particularly eager to divulge that information.
“Too often, a video offers an exciting glimpse of the future, but the robot isn’t available,” co-founder Charlie Kemp says in a release. “Stretch 3 is not vaporware. It is available today. It’s an invitation to join an amazing community creating an inspiring future. It’s also the most fun I’ve ever had as a developer.”
All of this is true — save, perhaps, for the last bit. We’ll just have to take the good doctor’s word for it. But just because it’s on sale today doesn’t mean most people will – or should – buy it. Like the Nvidia example above, it’s more properly seen as a reference device that third-party developers can access to make the kinds of apps that could – one day – be really useful.
Back to the question posed at the beginning. Why did we wait so long for a proper Roomba sequel? This product was designed to do one thing competently and has gotten much better at that job over time. The original Roomba had a hockey puck design, and it honestly hasn’t strayed far from the first on that front. There are extreme limitations to this form factor, however, including height (this matters a lot when it comes to where the embedded sensors are placed) and a lack of bezels.
Image Credits: Hello Robotics
As for this second part, Hello specifically refers to the recent excitement around humanoid robots. The concept of “general purpose” comes up a lot. Remember, for example, when the Tesla Bot was first announced and the company’s CEO promised a robot that could work all day at the car factory and then grab you some groceries on the way home?
It would take a lot more words than I’m devoting myself to right now here to explain why truly generalized robots are a lot further away than you think. I’ve often discussed a middle ground between the two – moving from single-use robots to multi-purpose robots. The path there may indeed involve an SDK and an app store type approach to introducing new features.
In this case, one begins to ask the reasonable question how much does the next robot in the home need to look like us? The really compelling argument here is the stairs, but we’re a long way from the point where such mechatronic complexities can be delivered to home users at reasonable prices.
I find this bit from Hello’s press materials particularly interesting: “Hello Robot has pioneered a middle ground between simple, single-use robots and complex humanoid robots, showing that robots don’t need to be humanoid to perform a wide variety of exciting jobs at home.”
Mobile manipulation is a huge, huge hurdle to developing a proper home robot. Most likely the solution will be little more than a couple of hands glued to a Roomba. Instead of jumping right into building another robot in our own image, Stretch offers an operator that’s more in line with what I’ve seen from homegrown robot research projects like those found at the Toyota Research Institute.
I’d say, at the very least, this is a space worth watching, even though you’ll have to keep waiting patiently for your next robot friend.
