Nov 21 2022

Artificial Muscles

There are some situations in which biology is still vastly superior to any artificial technology. Think about muscles. They are actually quite amazing. They can rapdily contract with significant force and then immediately relax. They can also vary their contraction strength smoothly along a wide continuum. Further, they are soft and silent. No machine can come close to their functionality.

In engineering parlance, a muscle is an actuator – a component that causes part of the machine to move. Boston Dynamics has produced some impressive results using standard actuators, but even their robots’ movements tend to be, well, robotic – a bit jerky and stilted. Compare that to the movements of a jaguar, for example. Engineers have been working on developing muscle-like actuators for years, with some progress but far from ultimate success.

One of the properties of a biological muscle is called the force-velocity relationship – the faster the muscle fibers contract the more power they produce. A second is the force-length relationship, essentially the longer the muscle the more power it creates. As a recent study points out:

However, it still remains a challenge to realize both intrinsic muscle-like force-velocity and force-length properties in one single actuator simultaneously.

In addition to these properties, to be more muscle-like we would need an actuator that can smoothly vary its power and also have soft components. There are other important properties, such as intrinsic response to load (does the system react to a load by contracting), static force (maintaining a load without moving), and the strength of the material used (how much of a strain can it take). Researchers, therefore, have been essentially trying to duplicate the structure and function of actual muscle to achieve all these properties. In the above study, for example:

This study presents a bioinspired soft actuator, named HimiSK (highly imitating skeletal muscle), designed by spatially arranging a set of synergistically contractile units in a flexible matrix similar to skeletal musculature. We have demonstrated that the actuator presents both intrinsic force-velocity and force-length characteristics that are very close to biological muscle with inherent self-stability and robustness in response to external perturbations.

Basically, artificial muscle. However, these systems are still relatively weak, which limits their applications. The HimiSK system uses material that changes shape under some external stimulus. An older system, the McKibben actuator, uses pneumatic pressure.

Another approach to soft muscle-like actuators is the electrostatic actuator. These use the attractive or repulsive force of electrostatic charge (the Coulomb force) to move a charged body. The problem with this approach is that high voltages are necessary to produce sufficient force to be useful. So again, these systems tend to be weak and therefore used for only small objects. Higher voltages are dangerous and can be harmful to the system and things that it interacts with. The goal of researchers is to find material that can produce greater force at lower voltages. A recent study reports doing just that: Lowering of Electrostatic Actuator Driving Voltage and Increasing Generated Force Using Spontaneous Polarization of Ferroelectric Nematic Liquid Crystals.

The study was published by Professor Suzushi Nishimura and his team from Tokyo Tech. Apparently the innovation is in switching from paraelectric to ferroelectric materials. A paraelectric material can have an induced electrical polarization under an external electric field (just as a paramagnetic material become magnetic under an external applied magnetic field). By comparison, a ferroelectric material will maintain its polarization even after the external electric field is removed. This provide two advantages for electrostatic actuators. The first is that lower voltage can be used to build up polarization over time, rather than applying a high voltage field all at once.

Second, the tested ferromagnetic material has a linear relationship with polarization and resulting generated force. This meets one of the muscle-like criteria – a smooth continuous alteration in produced force. Further, the force produced is 1200 times higher than with paraelectric materials. The resulting contraction of the material was 16% of the original length, another important factor which determines how much movement can result from an actuator of a specific size.

Whether or not this advance will make electrostatic actuators the leader of the soft muscle-like actuator race remains to be seen. It’s a good thing that there are multiple parallel research projects looking to solve this problem with different approaches. We’ll see which one prevails, or if the varying approaches will all have their preferred applications.

The ultimate goal, however, is to have a soft muscle-like actuator with all the properties of living muscle, or even better. That is still a high bar. Along the way there will be lots of applications of such actuators. However, there are some specific applications of true muscle-like actuators once we cross that line. The most obvious is for replacement limbs. It always amuses me when sci-fi shows pull back the skin on a very humanlike android robot to reveal nothing but hard metal beneath. Even Luke Skywalker’s hand had metal actuators. I know this is the movie cliche, the language used to communicate efficiently with the audience. In one glance they need to know they are looking at a robot. But this is highly unrealistic. Advanced robotic limbs or full androids would almost certainly have soft actuators that closely mimic the function of living muscles.¬†¬†Such artificial muscles would allow for human-like control, with soft parts that are compatible with the human environment. They will also be relatively silent, like living muscles, rather than making that characteristic actuator sound (like a droid).

Soft muscle-like actuators would also help get the Boston Dynamic robots to the next level, truly mimicking the function of living animals. Perhaps they will use a combination of traditional and soft actuators, to provide strength for some applications with finesse and grace for others.

Another critical feature for applications of soft actuators is their energy efficiency. How long can the onboard battery power a limb or an entire robot? Efficient muscles will be critical for many applications. Imagine a fully functional robotic artificial limb, but the battery only lasts 20 minutes. Right now the robotic dog can last about 90 minutes on their battery. This is good, but may be limiting for many potential applications.

I consider soft muscle-like actuators to be one of those stealth technologies – it is advancing in the background without much public attention. But at some point the technology will cross a line, and suddenly they will be everywhere, opening up new technological possibilities.


No responses yet