Do you know this theory that expectations of an innovation evolve according to the same curve, right?
👉 https://fr.wikipedia.org/wiki/Cycle_du_hype
The humanoid robot is an innovation that illustrates this phenomenon particularly well.
This cycle generally begins with the media emergence of a generic innovation, resulting from basic research.
At this point, the technology is still immature, but it is strongly attracting attention.
It responds to a clear functional expectation, and often to a shared fantasy: the gift of ubiquity with IoT, the vision of the future with AI, infinite movement with nuclear fusion… or even the augmentation of humans through fitness, prosthetics, or exoskeletons.
Then comes the peak of expectations.
Promises are getting out of hand, projections are multiplying and innovation is perceived as imminent and transformative.
After this peak, disappointment almost always follows.
At that point, either real uses eventually emerge thanks to those who persist, or technology collapses mediatically. In some cases, it disappears temporarily before returning later in a new cycle, such as AI and, precisely, robotics, which has promised us Terminator since the 90s.
The humanoid robot at the start of the peak
Humanoid robotics, it’s us in doubles, in triples…
It is a direct projection of man into the machine.
It also places us in a role of Creators, which further strengthens the collective imagination.
All the ingredients of the hype are there.
In this context, the last editions of the CES have clearly reinforced this peak in expectations.
Humanoid robot : still limited capabilities
At present, however, many tests demonstrate the real low capacities of humanoid robots in complex situations.
They’ll probably get there eventually.
But in the meantime, one point is becoming increasingly clear:
The hype seems to benefit our friends in maintenance above all.
A mechatronic factory
Here is the rough math on these real mechatronic gas plants.
The average MTBF of a good brushless motor, which is a little stressed without being abrupt, is about 50,000 hours.
Taking an optimistic assumption, this corresponds to 20 years of operation at a rate of 6 hours per day.
This means that there is a high probability that a given actuator will fail within 20 years, if the robot has not been discarded beforehand.
In the best-case scenario, a motor will therefore fail approximately every 250 days, simply because a humanoid robot carries around thirty actuators.
For example, the Atlas robot from Boston Dynamics has 28.
If one adds to this the risks related to elastomer seals, electronics and batteries, a conclusion is inevitable.
👉 The business of the humanoid robot will be the gold of maintainers and the ruin of users.
Never forget the hype.
And always think about the uses.
