As I had to prepare my slides for the talk Ill give next week during Web2Day conference in Nantes, it was a great opportunity to finally sit down and gather my thoughts about hardware design in robotics.
Basically, Im saying that it’s impossible to have the robots we’ve been dreaming of at the moment. We’re still far from there.
This thing is hard to admit for manufacturers, because all they see is that they keep doing better and better, one generation after the other. And they are right; they do! But for the public, interacting with a humanoid social robot at the moment is most of the time a deceptive experience, because it fails to deliver up to the standards of our dreams. And the more the robot looks like a human, the higher the expectations, the bigger the gap with reality.
Im also stating that out of universities and research labs, there’s no real market yet for these robots. Except SoftBank in Japan, I don’t think there’s any example of larger deployment.
I get the feeling that humanoid robotics companies are created by people who think robot before thinking market. And once they have their robot (and for some of them after they realize it’s not ready for the consumer market), they start wondering what it could be useful for and which industry could buy it.
On the other hand, you got companies that start from studying a market to then create a robot solving a problem for this market. The best example is iRobot: they picked the vacuum cleaner market but they did not decide to create a robot looking like a human that could use a vacuum cleaner. They robotized the vacuum cleaner. Roombas are robots, even if you might not think of them as such.