Image: © Lialore (CC BY-SA 4.0)

Soft robotics isn’t new, but as Johns Hopkins University’s Professor David Gracias and Aishwarya Pantula explain, the day of gelatinous robots capable of locomotion without human interference has already arrived.


Profile: David Gracias

David Gracias is a Professor at the Johns Hopkins University with a primary appointment in the Whiting School of Engineering and secondary appointments in the Krieger School of Arts and Sciences and School of Medicine.

Profile: Aishwarya Pantula

Aishwarya Pantula, is a PhD candidate in the Department of Chemical and Biomolecular Engineering, was recently awarded ‘Outstanding Presentation by an Early Career Researcher’ by Physical Review Journals, published by the American Physical Society.

Pantula won the award for her presentation at the 2023 Gordon Research Seminar about gelatinous robots which move in a direction dictated by temperature changes and design.

When we think about robotics, all too often we imagine something made of a hard exterior, perhaps with a face painted on it to give the idea of humanising the complex electronics within. But not all robotics is about humanesque robots. AI self-drive cars, for instance, are robotic, as are the automatic vacuum cleaners that hoover our homes while we get on with something different.

But there are different areas of robotics and one, in particular, is gaining interest: robots that act without human interference, but perhaps use a combination of heat, light, wind, air pressure, friction and any number of other factors that may lead to robots that seemingly have their own intelligence without the need for microchips, wires or, indeed, direct human interference.

Drawn to this fascinating area, Aether caught up with Professor David Gracias and PhD student Aishwarya Pantula of the Gracias Lab at Johns Hopkins University to discuss their area of expertise: gelbots. Years of research has led to the development of a ‘gelbot worm’; a stimuli-responsive gel that looks like a worm and its natural capabilities of swelling and shrinking can be manipulated so that the worm appears to be moving on its own.

Image: © Lialore (CC BY-SA 4.0)

Aether: Can you give a little of the history of gelbot research?

I am reminded of a provocative statement made by a former professor, ‘Biology is water’. It is striking though that water is all around us and in abundance. The Earth’s surface is about 71% water covered, and our own bodies have a majority percentage of water. Gels are unique materials composed of a large water fraction and the term ‘hydrogel’ can be traced back to Thomas Graham’s use of the word to describe a swollen jelly solution in 1864. Quite simply gels can be thought of as a kind of three-dimensional molecular fishing net that can be filled with water and swell or de-swell (shrink) and contract on the removal of water. Since gels have a large water content, they are of interest to human life not just in cosmetics, but also for drug delivery and creating replacement tissues and organs.

Aether: More specifically, what is your history and experience with soft robotics and gelbots?

We have heard a lot about how robotics will transform our lives such as with self-driving cars. Many of the advances have been made in algorithms and software and it is long-envisioned that intelligent machines could embody smart features.

However, while looking at our robot vacuum or self-driving cars or flat touch screens at a supermarket checkout, we notice immediately that these are made of hard materials such as metals and plastics.

The reason for this is that humans have known how to build machines using these materials and we have had electromagnetic motors and sensors and batteries to power them for a while now. Yet, these machines don’t touch and feel like humans.

Enabling soft-touch in machines is a very important feature, especially for human collaboration but yet to be embodied in robotics.

Harry Harlow, the American psychologist’s experiments showed that infant monkeys preferred clinging to cloth (soft) mother surrogates with no food even when offered food by wire (hard) mother surrogates.

Thus, incorporating soft-materials in robots is an important step to create soft-machines that are more akin to living organisms. It is envisioned that such machines would be more familiar to us since they would touch and feel like the organisms, animals and even our own bodies.

So soft-robotics is a grand challenge in engineering and given my laboratory’s interest in materials and miniaturised and intelligent systems, we are naturally drawn towards addressing this challenge.

Image: © Aishwarya Pantula (Gracias Lab, Johns Hopkins University).

Aether: What set you down this specific path of research that led to the gelbot worm?

I am driven by curiosity about the world I live in and the amazing engineering feats that living organisms perform all around us.

I have spent sleepless nights wondering how Life originated and I have been deeply interested in how nature organises its many components and how it can sense and communicate using chemical and environmental signals.

One of the key challenges that I have become interested in recently is to uncover mechanisms that can create a kind of intelligence in materials in the absence of wires, microchips, or a central processing unit.

We have been working on stimuli-responsive gels for over a decade in my laboratory. These are gels that swell or shrink (often by over a factor of 10 x volume) in response to environmental or external stimuli such as temperature, pH, or even specific biomolecules such as DNA.

Usually, this swelling or shrinking occurs because of a change in the molecular configuration of the molecular backbone (i.e., 3D fishing net) such as a coil-to-globule transition causing a change in hydrophilicity or charge and consequently driving water to enter or exit.

While such transitions can power dramatic shape change in gels, swelling and shrinking are symmetrical, so if something swells and shrinks it stays in the same place. Thus, these cyclic transitions have limited applicability to locomotion.

One trick to get a robot to move that was previously utilised was to pattern the surface with an asymmetric ratchet and in that case the symmetrical swelling and shrinking would cause the robot to move on the asymmetric surface since the surface broke the symmetry.

However, that severely constrains the motion of the robot; ratchets are not readily available. So could one design intelligence in the robot itself so that it could move unidirectionally on a ‘flat’ surface.

This is the key breakthrough that we achieved here. The PhD student in my lab, Aishwarya, systematically varied the asymmetry so that the robot performed a kind of breakdance move ‘the worm’.

Despite the surface being flat the undulation of the gel robot caused a variation in the contact area along the robot body. Since the shape and morphology was asymmetric, we could design the robot to move either left or right based.

Aether: The gelbot worm design appears very intricate - can you say a little about the design process that led to the 3Dprinted prototype?

The gelbot has three components. Bilayer segments on two sides connected by a suspended linker. The suspended linker is important to ensure that the contact area of the segments is transmitted into motion.

While the designs look complex, we discovered a simple rule that the robots moved consistently in the direction of the bilayers with higher contact forces.

Moreover, the magnitude of displacement could be tuned by tuning the dimensions of the robot.

Image: © Aishwarya Pantula (Gracias Lab, Johns Hopkins University).

Aether: You mention several possible future applications of this gelbot worm. Can you go into a little more detail about your hopes for a) delivering targeted medicines within the human body, b) how you hope to get them to react to human biomarkers and biochemicals, c) how and for what purpose do you see them as ‘patrolling and monitoring’ the ocean surface, and finally d) how far away are we potentially from seeing these gelbots fitted with cameras and sensors, and in which areas do you see this as having potential?

It is important to note that we created gelbots that move autonomously in response to cycles of heating and cooling. We note that temperature and chemical cycles exist all around us. For example, it is typically cooler at night than during the day. Our blood sugar levels rise and fall. So we envision that using this principle of using cyclic variations in environmental cues to drive swelling and shrinkage of gels to realise soft-robot locomotion across scales.

We envision using gelbots much like we might use drones or matchbox cars. They could move around, in response to cyclical variations of environmental stimuli such as biomarkers, dispense drugs or image the environment. They would be soft much like worms and be powered and communicate by similar principles that govern our motion.

I should caution we are still quite far away from achieving this vision, but the pieces required to achieve this path are being rapidly developed. Our demonstration of how ‘intelligence’ can be incorporated into shape and morphology is an important step in this direction.


David Gracias and Aishwarya Pantula


Gracias Lab

Johns Hopkins University