“If we want truly intelligent robots, we need to improve the design of robot bodies”

Why is there a need for an alternative to conventional robots, Mr. Milana?

Edoardo Milana: Of course, robots can already perform amazing things today with the help of artificial intelligence and machine learning. However, all this intelligence focuses on the software - the brain - and there is no comparable focus on the mechanical design - the body. So, robots are somewhat like puppets. Software is used to try to exert full control over all movements of the body. This approach assumes that the hardware is mechanically very simple and can be operated more easily with digital microcontrollers. Depending on the application, this may be sufficient and even necessary to meet the requirements for precision and the exertion of large forces. However, when considering movement efficiency and agility, the performance of robots lags far behind that of living beings.

There are already robots that imitate animals like dogs or cats.

Milana: These quadrupeds - and even humanoids - are indeed impressive technical masterpieces, but they cannot match real animals in terms of agility. They also consume a lot of energy to move, while animals and humans can perform much more complex movements with much less energy consumption. A quadruped running at a normal pace consumes about 300 watts on average to power its twelve motors, the 'muscles' of the robot, while a dog uses 30 watts to activate hundreds of muscles. This is possible because movements in nature depend heavily on the mechanical properties of the body. It adapts passively and actively to the forces exerted by the environment by taking advantage of the compliance of biological materials. In addition to digital control, the focus in robotics should also be on implementing intelligence or 'embodied intelligence' into the design of the robot. This would free up computing capacity and energy currently used for low-level motion control for the robot's higher-level logical operations, such as thinking, planning, and perceiving.

The term embodied intelligence originally comes from philosophy and psychology. What does it mean to you as an engineer developing robots?

Milana: What interests me is that the theory behind it can be applied not only to biological beings but also to robots. The basic idea of this theory is that the physical interaction between the body and the environment brings about intelligent behavior. It's not just about having a body controlled by the mind; this control is partly in the body itself and in the way it interacts with the mind. In robotics, this means: If we want a truly intelligent robot, we can't just take a body consisting of two or three metal rods and a few joints and then put a very intelligent computer inside. If that were the case, we would already have robots with completely different capabilities.

What might such intelligent robots look like instead?

Milana: I research soft robots made of soft materials, which could be considered inspired by primitive and aquatic biological organisms. There are already robots in this field whose control is entirely based on physical principles and therefore do not require digital microcontrollers. They use the nonlinear physical properties of soft materials to generate the control signals that drive the robot. Together with researchers from Stuttgart, the Netherlands, and Belgium, I have authored a study introducing such soft robots that introduce a new concept: the concept of 'physical control.' We have identified three special control mechanisms for such soft robots. An interesting example is robots with self-oscillating valves. When air pressure is applied, the valves open and then close again, increasing and then releasing the air pressure. This transmits a rhythmic air pressure signal through the system that controls the movement of the individual robot parts. In the future, we will have to find a compromise: We will not be able to do without software and microcontrollers in robotics, but we can achieve a lot through better design of the robot bodies.