Engineering news

How to control your soft robot

Tanya Blake

A tiny, soft, snake-like object is slowly but steadily moving inside a patient, in a sparkling-clean operating theatre. It’s not a snake though – it’s a robot.

A steady hand of a doctor is controlling the machine, with magnets held just above the person on the operating table. When the snake-robot is at the right spot, the doctor uses the magnets to release vital drugs right where they are needed.

We are not totally there yet, but the above is not all science fiction - researchers around the world have been developing such soft robots for the past several years. In the near future, they will be able to advance surgery and medical treatments, while other kinds of autonomous soft robot designs can explore the real world, safely, close to humans, in a way that traditional metal robots cannot.

However, researchers still need to solve a couple of big challenges: how best to power and control these squidgy machines.

In March 2017, engineers have made a big step forward in controlling soft robots - by using magnetic fields to remotely move iron chains made up of tiny particles embedded in the machines.

Joe Tracy, an engineer from North Carolina (NC) State University and the lead author of the paper describing the research, says that using “self-assembling” chains that consist of iron microparticles means they can get the simple robots to perform more complex functions. This could one day lead to soft robots used as remotely triggered pumps for drug delivery or “robot structures that can be remotely deployed when needed” for things like disaster aid, adds Tracy.

The researchers introduced iron microparticles into a liquid polymer mixture and then applied a magnetic field to make them form parallel chains. The mixture was then dried, leaving behind an elastic polymer thin film embedded with the aligned chains of magnetic particles.

They found that by varying the direction and strength of the magnetic field the chains of microparticles in the robots would respond by moving themselves and the surrounding polymer in the same direction.   

Photo of soft robot chains controlled by magnets. Credit: North Carolina State University

Obstacles ahead

Among the most important challenges in soft robotics are actuators and control, says Cecilia Laschi, soft robotics researcher at the Biobrobotics Institute, Pisa, Italy. She says that the idea of controlling movement of soft robots by an external magnetic field is “very interesting” and could help to create all kinds of external controllers.

Of course, she adds, the controllability of these robots and the magnetic field needs to be explored further to see if they could work well in real-world environments. Because the approach taken by the NC State University requires a person to control the external magnetic field close to the machine, other methods would need to be considered for autonomous soft robots, says Laschi.

The researchers at NC State University have created three kinds of soft robots that, embedded with iron microparticles, can be controlled with magnetic fields.

One device is a cantilever (an arm or beam structure that is fixed at one end) that can lift up to 50 times its own weight. Another system is an accordion-like structure that expands and contracts, mimicking the behaviour of muscle. The third is a tube that is designed to function as a peristaltic pump – a compressed section travels down the length of the tube, much like someone squeezing out the last bit of toothpaste by running their finger along it.

"We're now working to improve both the control and the power of these devices, to advance the potential of soft robotics," Tracy says.

Laschi says that the idea of controlling the movement of a robot externally to the machine but right next to it is “very suitable for applications where the robot is doing an exploration inside something”.

Aerial soft robotics expert Sina Sareh, from Imperial College London, agrees that while this method of actuation can be useful for a wide range of robotic applications, it will be especially useful for the "next generation of medical instruments that can be constructed with a lower profile, yet able to generate required levels of torque and preserve certain levels of flexibility".


Control solutions

Controlling medical devices using magnetic fields is not new and has been explored by a number of researchers. For example, teams at Vanderbilt University in the US, the University of Leeds in the UK, and in Scuola Superiore Sant'Anna in Pisa, Italy, all looked at steering and moving capsule endoscopes. Researchers at the Chinese University of Hong Kong have also magnetically steered micro-robotic swarms.

However, Helge Wurdemann, head of the Soft Haptics Lab at the University College London, who was not involved in the North Carolina State University research, says the approach is still considered novel in soft robotics. By introducing the iron micro-particles into a liquid polymer mixture, he explains that the “holistic, inherent softness of the robot is preserved”.

At the same time, Wurdemann adds that it is “fascinating that one of their robots can exert high forces” and lift up to 50 times its own weight. The research community has instead been exploring “stiffening principles” to produce similar results for autonomous soft robots.

Typically, an autonomous robot must have a basic body structure, sensors, a central control system (microprocessor), actuators (motors), a power supply and a programme for its behavior. Building a body from soft materials, like polymer, can be done by casting, injection moulding and 3D printing.

Sensors and microprocessors can now be manufactured small enough to be embedded inside a soft robot without compromising its flexibility. Electronic components too can be made to be fairly flexible or even stretchable. However, traditional electric motors can’t be shrunk down and embedded in the same way that sensors can, plus they become less powerful the more you shrink them.

Hydraulic and pneumatic systems have been used to control soft robots but they have to be tethered to the machine – not so useful if you want a robot to travel long distances. Other options have been to use so-called soft actuators such as electro-active polymers (EAPs), macroporous gels, and other phase-transition materials, but much more research needs to be done before they are effective motors.

Some of the most widely used ‘soft’ actuators use threads of miniature shape memory alloy wires or foils. These are alloys that change their shape when they are heated, acting similar to muscles. So far, they have been found to be fairly inconsistent, energy inefficient and easily affected by environmental conditions.

Some of Wurdemann’s own research focusses on surgical robotic soft manipulators, looking at ways to change between soft and stiff states. This is in a bid to give soft robots more control over movements – making an arm fluidly move between a soft and rigid state when needed for different tasks.

With colleagues, Wurdemann applied an ‘antagonistic actuation’ principle to soft silicone-based robots that takes inspiration from the octopus. The octopus activates two sets of “collaborating” muscles allowing the animal to stiffen its arms. The approach uses pneumatic and tendon-driven actuation in robotic soft manipulators –  a mechanism commonly found in animals where muscles can oppose each other to vary joint stiffness.

The approach was said to have gone beyond what state-of-the-art soft, flexible robots can achieve when the research was published in 2014. This was because the robot, mainly made of thin fabric-like materials, is filled with air to achieve a fully-extended state, it can be shrunk to a considerably small size when entirely deflated. This allowed for a wide range of actions to be carried out on-demand, useful for minimally invasive surgery. It also opens the door for highly adaptable robots useful for search and rescue missions.

Batteries not included

But making sure these soft robots have enough power to move outdoors for hours upon end is not an easy task. Anything that is electrically powered must store energy in batteries or capacitors. Although these can be made relatively flexible they are not widely commercially available and cannot store large amounts of energy. Researchers are instead turning to a biologically inspired solution: storing chemical energy in the soft robot and using it when needed just like fat is transformed into sugars, fats and proteins to provide energy for migrating birds.

Wendelin Jan Stark, professor of chemical and bioengineering at ETH Zürich, Switzerland, says that while the work done by NC State University is interesting, that using magnetic field driven soft robots is ultimately constrained. It would take incredibly large amounts of magnetic power to move larger soft robots out in the real world.

Instead, Stark and his team at ETH Zürich are exploring the use of chemical reactions and pressure driven movement into silicone robots. This too is how Harvard University researchers managed to power a 3D-printed ‘octobot’ that has no electronics. They announced the ‘world’s first’ autonomous, untethered, entirely soft robot in August 2016.

They were able to 3D print each of the components required within the soft robot body, including the fuel storage, power and actuation and control systems.

Photo of the Octobot. Credit: Lori Sanders, Ryan Truby, Michael Wehner, Robert Wood, and Jennifer Lewis

Harvard’s octobot is pneumatic-based powered by gas under pressure. A reaction inside the bot transforms a small amount of liquid fuel (hydrogen peroxide) into a large amount of gas, which flows into the octobot’s arms and inflates them like a balloon. To control the reaction, the team used a microfluidic logic circuit based on work by co-author and chemist George Whitesides. The circuit, a soft analogue of a simple electronic oscillator, controls when hydrogen peroxide decomposes to gas in the octobot.

Robert Wood, who led the research, says it demonstrates that “we can easily manufacture the key components of a simple, entirely soft robot, which lays the foundation for more complex designs”.

Stark believes that when it comes to the recent NC State University research automisation and the industrial robotics sector will profit first, then the biomedical industry, when it goes towards “enhanced biomechanics, for example, novel movement assisting devices”. Repetitive, soft action is also of particular interest, he says, in the movement of fragile goods but the magnetic field needed would be “large, energy consuming, and so perhaps less useful”.

However, when it comes to autonomous soft robots controlled using the perhaps more promising approach of chemical reactions, the applications could be much wider. Robots with the muscle control and dexterity of animals and humans could be used for handling delicate items in warehouses or creating powerful exoskeletons for the elderly.

The technology could even be used to create homes that could morph or change depending on its environment, like a soft robotic wall that could morph into different shapes when needed.

Although it may sound far-fetched, with the field of soft robotics making strides every day, it shouldn’t be long before we start to see these ideas become a reality.  



Read more related articles

PE Magazine

Current Issue: October 2017

  • Holly's comet
  • Batteries not included
  • Printing buildings
  • North Sea decommissioning
  • The end of SOFIA?

View all

PE app

  • Industry features and content
  • Engineering and Institution news
  • News and features exclusive to app users

Download the PE app

PE newsletter

A weekly round-up of the most popular and topical stories featured on our website, so you won't miss anything

Subscribe to the PE newsletter

Opt into your industry sector newsletter

Related articles