The industry’s main assumption regarding humanoid robots is that the best way to make money in the near future is by working in warehouses or factories. It’s not hard to understand where this assumption came from: repetitive tasks that require strength or flexibility, in well-structured settings, are one place where robots seem to thrive. And if you want to make billions (because your company is valued at that amount), there don’t seem to be many other options. Cartwheel Robotics (19659002)is trying something different with humanoids. Cartwheel Robotics is more interested in creating robots people can connect with. The ultimate goal is to create general-purpose home companionship. Cartwheel founder Scott LaValley describes Cartwheel’s robot as a “small, friendly humanoid designed to bring joy warmth and a little bit of everyday magic to the spaces we inhabit.” It’s expressive and emotional intelligent–not just technology, but a presence that you can feel.

The rendering below shows the design and size of Cartwheel’s prototype humanoid. Cartwheel

Making a commercially viable robot has always been a challenge. In less than a decade, a number of social home robots, backed by substantial investment, tried to convince consumers of their worth. They failedCartwheel’s decision to go the humanoid route with legs and all is a sign that they are not sure if the fundamental problems of social home robotics (cost and novelty) have been resolved. This means that Cartwheel has to deal with a variety of issues, from motion planning to safety to balancing. All this while making sure the robot is reliable enough to be used around children.

LaValley may be one of the few people capable of making a commercially viable social humanoid. He spent nearly a decade working on Atlas robots at Boston Dynamics, and then five years at Disney where he was the leader of the team that created Disney’s Baby Grootrobot.

Building Robots To Be People’s Friends.

In terms of humanoid robotics, there’s a big difference between the versions Atlas that LaValley has worked on (DRC Atlas specifically) and Baby Groot. LaValley says he was most impressed by his children’s reaction when he showed them the robots that he was building. LaValley recalls that Boston Dynamics was known for its terrifying robots. “I was excited to be working on the Atlas robots, because they were cool tech, but my children would look at them with a huge smile and ask, “Is that Baby Groot?” At Disney, my kids would light up and ask, “Is that Baby Groot?” LaValley’s vision of robots is quite the opposite of the current generation of commercial humans. You could say that they are designed to be work-oriented and not to be friends, but the majority of design choices are based on what will be most appealing to the public, as well as investors, in a boringly “futuristic way. There are many reasons to design a humanoid that has commercial (or industrial) aspirations in a certain manner. But for better or for worse, no one is going to like those robots. Respect them? Sure. Do you think they’re cool or not? Probably. Want to be their friend? Not likely. LaValley says Cartwheel has a great opportunity. “These humanoid robotic tools are designed to be tools. They are soulless. They’re soulless. Cartwheel’s robots may need to be practical in the future (as shown in this rendering) to find a home in people’s houses. Cartwheel

Yogi, one of Cartwheel prototypes. LaValley describes it as having “toddler-like proportions,” that are key to making the robot appear friendly and approachable. It has rounded edges, a big head and is even a bit chubby. I don’t think of Yogi as a robot; I see him as a character. The second prototype, Speedy, is less complex and intended to be a more customizable commercial platform in the near future. Imagine Baby Groot but available in any character and to companies other than Disney. LaValley informs us that Speedy will be accompanied by a special torso for a “particular outfit” in the near future.

The previous generation of social robotics learned the hard lesson that it takes more than just good looks to make a robot connect with humans on a long-term basis. LaValley believes that AI is one solution to this problem, as it could offer a way to preserve novelty by keeping the interactions fresh. Cartwheel is also experimenting with AI for whole-body movement generation, where robot behaviors will be unique even when given the exact same inputs or conditions. Cartwheel’s Home Robots Plan (19659014)

Cartwheel has started with a commercial platform but the ultimate goal is to bring these small social humanoids home. It is important to consider safety and affordability differently when designing humanoids for warehouses and factories. Cartwheel’s small robots will help with both, but it’s still a robot which is likely to cost quite a bit. It’s more than an appliance, though not as much a car. That’s the amount LaValley is willing to commit at this time.With such a price come high expectations. For most people, buying a humanoid for the home will only be justified if it is both practical and lovable.

LaValley admits that he doesn’t know all the answers. “There’s lots to figure out.” One increasingly popular approach with robots is the service model. The robot is essentially rented, much like you would pay for a gardener or housekeeper. Cartwheel’s robots must be able to justify their existence financially for this to make sense. LaValley says that the problem will not be solved in a year or even five years. “There are many things we don’t know–this will take some time.” Cartwheel

Cartwheel is a three-year-old company that started by offering robotics engineering services for corporate clients. LaValley was able to bootstrap Cartwheel’s robot development with the help of an initial funding round and is now delivering a dozen variations of Speedy to museums and science centres over the next year.

But the dream is to have small home robots which are both companionable, and capable. LaValley even uses terms like “general-purpose” to describe them. I see all robots moving towards general purpose. Our strategy isn’t to get general purpose or into the home on day one. But we’re working toward that goal. “That’s our north-star.”