[I'MTech] What is bio-inspiration?

How long has bio-inspiration been around?

Frédéric Boyer : There have always been exchanges between nature and fundamental and engineering sciences. For instance, Alessandro Volta used electric fish such as electric rays as inspiration to develop the first batteries. But it’s an approach that has become more systematic and intentional since the 1990s.

How does bio-inspiration contribute to the development of new technologies?

FB : In the field of robotics, the dream has always been to make an autonomous robot, that can interact appropriately with an unfamiliar environment, without putting itself or those around it in danger. In robotics we don’t really talk about intelligence – what we’re interested in is autonomy, and that’s still a long way off. 

There’s a real paradigm shift underway. For a long time, intelligence was considered to be  computing power. Using measurements made by their various sensors, robots had to reconstruct their complex environment in a symbolic way and make decisions based on this information. Through this approach, we built machines that were extremely complex but with little autonomy or ability to adapt to different environments. Through the bio-inspiration movement, in particular, intelligence has also come to be defined in terms of the autonomy it brings to a system.

What is the principle behind a bio-inspired robot?

FB : Bio-inspired robots are not based on the perception and complex representation of their environment. They are simply sensors and local loops that enable them to move in different environments. This type of intelligence comes from observing animals’ bodies: it’s what we call embodied intelligence. This intelligence, encoded in the body and morphology of living organisms has been developed over the course of evolution: an animal with a very simple nervous system can interact very effectively with its environment. We practice embodied intelligence every day: with very low  levels of cognition we solve complex problems related to the autonomy of our body.

To illustrate the difference between the paradigms we can take the example of a robot that creeps along like a snake. There are two approaches for piloting this system. The first is to place a supercomputer inside the robot, that will send a signal to each of the vertebrae and motors to drive the joints. With the second approach, there is no centralized computer. The “head” sends an  impulsion to the first vertebrae which spreads to the next one, followed by the one after that and is automatically synchronized through feedback phenomena between the vertebrae and with the environment, through sensors.

When robotics is inspired by nature

What if nature allowed researchers to reach the Holy Grail of robotics, namely an autonomous robot capable of adapting to its environment? This is the approach taken by Frédéric Boyer.

Listen to the podcast: (french)