Soft polymers for human-robot interaction

Most people know robots either from science fiction films or from company production halls. There, they have a programmed sequence of movements and reliably perform their tasks. But: they are heavy and rigid – anyone who gets in the way of their fixed motion sequences is in the way. For this reason, robots need to be developed that have a softer surface and are thus better suited for interaction with humans.

Junior professor Gianluca Rizzello and his team in the "Adaptive Polymer-based Systems (APS)" working group are developing intelligent polymer systems that can later be used in more sentient robots. In an interview with K-MAG, Rizzello talks about the development of these polymer systems, what makes them tick and how they can be used in the future.

Mr. Rizzello, why are new, intelligent polymer systems needed?

Gianluca Rizzello: Nowadays, actuator technology mostly relies on electric motors or pneumatic or hydraulic drives. Even though those systems are well established in the engineering practice, they are characterized by bulky and complex components like compressors and gearboxes, which might be unsuitable for some applications, for example in which lightweight and energy efficiency represent the primary requirements. Smart polymer systems represent a means to overcome those limitations, by providing a novel family of lightweight and multifunctional mechatronic systems capable of performance not achievable with standard actuator technologies. Some examples include energy-efficient pumps and valves, soft robots for safe cooperation with humans, wearable sensors and interfaces, and flexible loudspeakers which can be attached on curved surfaces.

What are dielectric elastomers and what characterises them?

Rizzello: Dielectric elastomers consist of rubber-like material which expand when subject to an electric voltage. A unique feature of dielectric elastomers is the ability of undergoing very large deformations of 100 percent and higher. In addition, those materials exhibit a feature referred to as self-sensing, which means that we can use them as actuators and sensors at the same time. Large deformations and self-sensing, together with lightweight, energy efficiency, high speed, and low cost, make dielectric elastomer a unique type of actuator technology.

Why are these elastomers particularly suitable for the development of sentient robotic arms?

Rizzello: The intrinsic softness of dielectric elastomers makes the resulting robot also safer to operate around people, thus allowing to implement modern human-robot cooperation paradigms. It is remarked, however, that the key for developing intelligent polymer robots is represented by self-sensing. In fact, only if we combine polymeric robots with intelligent algorithms – like a smart brain –, which are capable of properly interpret the 'nerve' signals and transform them into 'muscle' commands, we can truly exploit the intrinsic intelligence of such systems. This paradigm represents the key to develop the novel generation of novel smart, safe, and autonomous robotic technologies.

Where else could the polymers be used in the future? What applications are conceivable?

Rizzello: Smart polymer systems can be practically used in all mechatronic applications in which high energy efficiency, lightweight, and safety are required. Novel soft robots and wearable technologies represent primary examples where those requirements appear as fundamental, of course. At the same time, dielectric elastomers can be oftentimes used to replace current actuator technologies, such as motor, with equivalent artificial muscles that provide more energy efficiency without reducing the dynamic performance. As an example, me and my research group have worked in the past to replace conventional magnetic valves and pumps with equivalent dielectric elastomer systems, obtaining energy savings up to 99 percent. This result shows how polymeric actuators can be potentially used as key component to develop novel and more sustainable technologies of the future.

How do you think research into intelligent polymer systems will develop in the future?

Rizzello: Over the last two decades, significant research has been conducted on dielectric elastomers. On the one hand, material scientists have developed more robust and reliable materials, which are suitable to be used in real-life environments. On the other hand, engineers have used those materials to develop a large number of mechatronic systems, such as actuators or robots. What is really missing at the moment is the availability of intelligent algorithms capable of driving them efficiently and autonomously. Few researchers, including me and my group, have already developed some of those algorithms for very simple types of dielectric elastomer systems, like one-dimensional valves. Development of AI algorithms capable of dealing with more complex systems, such as robots driven by dozens of artificial polymer muscles, is still an open research issue.  Only once we properly understand how to design and control such types of systems, we will be able to truly exploit the intrinsic intelligent of smart polymer systems.