Jinzou Ningen – Artificial Human project


Exploring possibilities

Making Things Move: Muscles As A Fundamental Requirement

Whether we’re talking biological lifeforms or robotic creations, there is one fundamental property they all share: the ability to displace themselves. Bacteria and similar propel themselves forward using flagellum, thin, whip-like structures, directly driven using molecular motors. Larger, multi-celled creatures begin to use muscle-like structures. Robots use electric motors, steppers, servos and sometimes pneumatic artificial muscles [1] or electroactive polymers [2].

For a system like that of the mammalian body including that of humans, there is the option to use a rotational actuator, such as an electric motor or equivalent, implanted in the joints so as to directly move the joint. The major disadvantage of this is that it takes a lot more energy to move the joint this way than to use the approach used in biological systems: the very familiar lever system. As the muscle is attached on two points a fair distance on opposing from the joint, the exerted force when the muscle contracts is multiplied by the distance from the joint, effectively making it much stronger. The only disadvantage is that the muscle has to work over a long distance to generate this force. With pneumatic and electroactive polymer muscles this actuator system is emulated.

While many robots feature wheels and thus are fine with non-linear actuators, the use of wheels and propulsion methods based upon these have a lot of disadvantages, such as navigating non-flat terrain, ascending and descending stairs and ultimately a lack of sheer speed and versatility. For any universal body, the type used by primates including humans has proven to be the most versatile and universal, as its ability to interact with its environment is unparalleled. This is the body type I would pick for an advanced artificial human project.

Moving on, we have established that linear actuators akin to muscles are a definite requirement for advanced movement and maneuvering. We have mentioned pneumatic and electroactive polymer muscles as possible options. We will look at these two options as an alternative for biological muscles in our theoretical artificial human body.

Pneumatic artificial muscles (PAMs) are the oldest form of linear actuator emulating the function of the biological muscle (BM). Both shorten in length when activated, resulting in a pulling force across a certain distance. Main difference is that in a BM this shortening force is generated by the many myofibrils, which form the fundamental units of muscle fibers. A myofibril consists out of long proteins including actin, myosin and titin. The basic organization is into thick and thin filaments, repeated along the length of the myofibril in sections called sarcomeres. During contraction the thin (actin/nebulin) and thick (myosin/titin) filaments slide along each other. This contraction is triggered through the active provision of energy to the myofibril and the contraction is reversed upon ceasing this energy provision (neutral state).

A PAM is essentially a bladder made out of rubber or similar material with an inflexible net-like material surrounding it. When air (pneumatic) or a fluid (hydraulic) is pumped into the bladder it will swell up, shortening the length of the bladder as it draws both ends together. As a BM replacement PAMs come pretty close, as their compliant behaviour and simplicity of construction makes them both safe and cheap. They are however also bulky due to the required valves and air compressor. As the elastic material of the bladder’s membrane is constantly stressed, failures often occur here, in the form of ruptured bladders.

Electroactive polymers (EAPs) are a much newer invention. They’re polymers which exhibit a change in size or shape when exposed to an electric field. There is a large number of very different EAPs, but we’ll just focus on the ones relevant to our theoretical body build. These are Dielectric EAPs and Ionic EAPs.

Dielectric EAPs (DE-EAPs) [4] are based on the electrostatic force [5] and is similar in construction to a capacitor [6]. It has two electrodes with an insulating elastomer film between them. Applying a voltage to the electrodes causes the Coulomb forces between both sides to move the electrodes together, a movement allowed by the flexible elastomer film sandwiched between both sides. This results in an elongated movement. As this configuration is a capacitor the elongation doesn’t reverse on its own when an external voltage is no longer applied. This makes it very power-efficient. As with all polymers durability is an issue, however, even if DE-EAPs are otherwise an almost perfect albeit inversed approximation of BMs.

The most practical ionic EAP is an ionic polymer-metal composite [7], yet they aren’t nearly as universal as DE-EAPs, displaying only a deflection actuation (bending), making it far less useful.

In the end it seems that biological systems are more practical and versatile here. Although the actual energy to force conversion of BMs is fairly low, their ruggedness and self-repairing qualities make them quite unmatched at this point. Our best bet may lie in a type of artificial muscle which actually mimics the functioning of BMs on a molecular level. Call them neo-biological muscles, if you wish. I would like to discuss this option in another article.

Until then,


[1] http://en.wikipedia.org/wiki/Pneumatic_artificial_muscle
[2] http://en.wikipedia.org/wiki/Electroactive_polymers
[3] http://en.wikipedia.org/wiki/Myofibril
[4] http://en.wikipedia.org/wiki/Dielectric_elastomers
[5] http://jnaudin.free.fr/lfpt/html/bubble.htm
[6] http://en.wikipedia.org/wiki/Capacitor
[7] http://en.wikipedia.org/wiki/Ionic_polymer-metal_composite


Filed under: Actuators, ,


Maya Posch: professional software engineer and game developer. Graphics artist and all-around science junky.