Well, while searching for dielectric elastomer actuators I can find that they normally have efficiencies around 80% or more (source page 38), they are cheap and easy to build.

But the more I search about them, the more I notice that I can't find projects more complex than proofs of concept and others uses, like microfluidic pumps.

Why is this lack of interest on actually using this class of actuators on more elaborated robots? What I'm missing?

Artificial muscles using dielectric elastomers

This is the only example that I could find (source).

  • 2
    $\begingroup$ Efficiency is one thing, but performance is another. Efficiency doesn't get jobs done. Kind of like how ion engines are very propellent efficient but their low impulse makes them not viable for most applications. How much power, displacement (or speed), or force/torque can these output for their size? It's possible that they just aren't very performant despite having high efficiency. $\endgroup$
    – DKNguyen
    Commented Dec 28, 2023 at 20:12
  • $\begingroup$ @DKNguyen It depends on the specific actuator, some have 20% or less of stroke, others have 300%, some can lift 50 times its own weight, others a 1000, there is also the possibility of twisting fibers of the material, like it is done to fishing line artificial muscles. They actuate as fast as you input energy on them. $\endgroup$
    – Fulano
    Commented Dec 28, 2023 at 20:15
  • 1
    $\begingroup$ Because this is the first application I had in mind: electronicdesign.com/technologies/analog/article/21213099/… This article does state that their power density and lifetime are rather low, even when compared to other more exotic actuators. $\endgroup$
    – DKNguyen
    Commented Dec 28, 2023 at 20:17
  • 1
    $\begingroup$ A lot of them are tested for power, efficiency, and even lifetime (in cycles), but end up suffering in control. Muscles deliver all of these and more. Magnetism sets the bar. $\endgroup$
    – Abel
    Commented Dec 28, 2023 at 22:39

1 Answer 1


Preface: I am not an expert in Dielectric Elastomeric Actuators, but as a mechanical engineer I operate a small prototyping business that brings new technologies to market. Hopefully this can give you a manufacturer's perspective into some of the difficulties that I see in developing and profiting from this technology.

Currently, Dielectric Elastomeric Actuators (DEAs) are likely not commercially available because of a combination of several of the following factors:

  1. DEAs have a dangerously high working voltage. Even on a 2022 improved 10nm MIT design they still show voltage ranges between 300V and 1800V. While these voltages can be made safe for consumer equipment such as in florescent lights, the engineering, fail-safes, certifications, and liabilities make for a high barrier to entry into the market. If the voltage can be brought below 48V (perhaps with thinner layers), the safety risk and electrical certification is much less stringent.
  2. DEAs require complex manufacturing including spin coating, vacuum bubble removal, nano meter level precision, thin graphene or carbon nanotube electrodes, unknown electrode application techniques, electrical connection to the electrodes, high-voltage sinusoidal IGBT driver circut, dielectric breakdown detection for safe failure, and lots of other details unknown at this time. As graphene and carbon nanotube production and applications mature there will be less manufacturing hurdles to overcome.
  3. DEAs require prestretch to prevent Electromechanical instability. Having two different displacements for a given voltage can cause the actuator to unpredictably move between the two creating very high accelerations and damage the elastomer and or electrode causing failure of the actuator. I did not find a satisfactory explanation of the reason for the initial force/voltage peak, but my understanding is that the elastomer chains are not initially lined up and lining them up ahead of time ensures that there is not a high force required for an initial displacement. While prestretching simple enough to do in theory, this is another mechanical engineering hurdle to overcome with every physical implementation of the actuator. In the future this hurdle could potentially be addressed with chemistry and cure processes to create directional cross-linking in the unloaded state.
  4. Already proven amplified piezoelectric actuators fill many of the niche uses for DEAs. Piezo materials suffer from a lot of similar issues such as high voltage and preload required. The big difference is that this technology has already been developed and many of the engineering hurdles have been overcome. While being "soft" may provide a sufficient niche in many future applications, currently an off-the-shelf piezo actuator fills almost any need for a DEA.
  5. In rapidly developing fields it is always important to consider that progress is slowed by patents. Both by the researchers taking time to acquire patents, everyone having to perform a detailed patent searches to know if they can participate, and everyone avoiding key patent claims that restrict the best methods. Once a viable path to profit is realized, then patents accelerate development by encouraging the investment of capital, but until that point they are antagonistic. 3d printing exploded in popularity and development only after key patents expired.
  6. Just speculating here, but a lot of the scientists that would be experts in further developing DEAs are being pulled towards developing other highly financially incentivized technologies such as new batteries, fuel cells, gas sensors, ultra capacitors and other leading edge electrode technologies.
  7. There are actually lots of competing soft actuator technologies that I was unaware of. These dilute the focus on advancing one technology like DEAs and may ultimately have superior characteristics for applications that are not shadowed by piezo actuators.


  1. The Elastomeric Actuators Lift Aerial Micro-Robot article DKNguyen posted in the comments.
  2. The bottom of the Elastomeric Actuators Lift Aerial Micro-Robot article also has a “Supporting Information” link that includes the technical details of how they were able to improve the performance.
  3. First article from the OP: Dielectric Elastomers for Actuation and Energy Harvesting
  4. Second article from the OP: Realizing the potential of dielectric elastomer artificial muscles
  5. Soft Robot Toolkit looks like an excellent resource for developing prototypes.
  6. Dielectric Elastomers Wikipedia Article
  7. Soft Robotics Wikipedia Article

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.