One week away until we lift our glasses to science!
To continue our series of Pint Of Science 2016 previews we spoke to Michelle Reeve (@michelleareeve) about her forthcoming talk “Pint of Spiders… with robots” at The Rugby Tavern, London on the 23/05. Although it might initially sound like the stuff of nightmares it is in fact extremely important research and we cannot wait to hear the whole talk next week!
Michelle was born and bred on the east coast of Norfolk, UK, and now lives in London after moving there to study for her undergraduate degree: BSc BioVeterinary Sciences at the Royal Veterinary College.
When we think of robots, often the first image to spring to mind is that of a humanoid. Perhaps made of metal, slightly clunky but able to move around more or less like a human. The first robots were indeed like this, and this stereotype has been emphasised by numerous film and literature depictions of robots both before and since. As technology has advanced, robots have become sleeker, more efficient, more specialist, and much more varied.
One of the ways in which we have improved legged robots is by turning to nature. Animals can move elegantly and efficiently over obstacles and difficult terrain with ease. They have benefitted from millions of years of evolution, so that today, they are brilliantly well-adapted to moving around within their environments. Increasingly, research teams comprised of roboticists, engineers and biologists come together, with the dual goal of learning how a particular animal or group of animals move, and using this knowledge to create an efficient, ‘bioinspired’ robot.
Despite this, our legged robots are still far from perfect. They are still are often unsteady, clumsy and energy-hungry. If they are damaged, a human needs to fix them. Depending on the purpose of the robot, this can be okay; if a research robot gets broken and needs a quick fix, nobody minds But if the robot is designed to enter dangerous places, you really don’t want to be sending an engineer in with it.
And this is where my research comes in. I’m the biologist in this interdisciplinary team, and I work on spider movement. Why spiders? Well, they can do exactly what we’d like our robots to do: when they damage themselves, they can just carry on. Spiders can actually self-amputate their legs, through a natural defence process known as autotomy. This can happen as a result of fighting during mating, being attacked by a predator, or even getting stuck in a problematic moult. Juveniles can often grow the leg back during their next moult cycle, but adults just have to live with it. And they cope remarkably well!
My work on the biomechanics of spider locomotion hopes to tease out some of the secrets of exactly how they move with missing legs. I’m looking at wolf spiders, a native UK species which run fast overground, just as we’d like our robots to be able to do. By filming them with a high speed camera, I study their movement over flat surfaces and rough terrain. In particular, I look at how the individual legs pair up or group together – and how this changes when a leg is lost. This data can suggest how the spider gait is controlled, and from this we can design a bioinspired control system for a new or existing legged robot.
This type of adaptive, bioinspired control system could have real benefit to robotics, and to people. Robots typically house a suite of sensors, and these could report if a leg were to become damaged, or broken off completely. This would cause the control system to change to the ‘optimal’ gait for a missing leg, as inspired by the spider. Of course, this is a very simplified description of how it would work, but the positive impact remains. Robots like this could enter dangerous territory, coping with damage without needing humans around to fix it. They could be mounted with cameras to search for survivors alongside the emergency services, or used in the armed forces to sweep enemy territory for threats before in sending soldiers. So, by studying the movements of the humble spider, we can begin to develop bioinspired legged robots could save human lives.
Tickets for Pint Of Science talks are selling fast, so get over to through their official website to grab some.
Mendeley is extremely excited to be partnering with Pint of Science for the second year running! This year, we are sponsoring “Atoms to Galaxies” events across the UK, and Mendeley API & Mendeley Data are co-sponsoring “Tech Me Out” events. Last year was a massive success, and we feel passionate about the Pint of Science mission to bring research to the public, and give a chance for academics to present their work. We hope to help grow the event so more people can hear about the vast and amazing research happening in our galaxy — and beyond.
Mendeley Blog 
This spider-robot looks awesome. I guess that many problems will be solved when we will be able to advance in biorobotics, or will we?
Do you think that finally we will have only purely mechanical robots, or biorobots will be a reality in the future?
From a developer’s point of view, I hope that we stick with purely mechanical, chemistry and biology is not my strong point.
Thanks for the comment, Oliver!
From my limited first-hand experience of robotics, I suspect that although advancing in biorobotics may solve some of the problems this particular type of robot is hoping to address, it will undoubtedly create some problems of its own. Assuming you’re talking about incorporating (living) biological material into traditional mechanical robotics (or vice versa), a major consideration is ethics. How do we know when the ‘biorobot’ is alive? What experiments or tasks will be ethical to run with it? Of course, it will be many years before we have to address these sorts of questions, but they’re worth thinking about.
One of the other uses of a ‘bioinspired’ robot, as described in the above post, is to run experiments that are unethical / impossible to run with live animals. Imagine a robot with movement inspired by a dog. It would be great to know how a dog’s mechanical and nervous system deals with, say, being pushed over – how does it stabilise itself? Of course we cannot and would not want to do this with real dogs. But you could do it with the dog-inspired robot, and make educated guesses as to how its response corresponds to that of the dog’s own biomechanics. These sorts of experiments would potentially be unethical on a biorobot, depending on its composition, or at least would have limitations.
Personally, I think we will see biorobotics in the future, purely because of scientists’ innate curiosity! But I think there will be a place for both mechanical robots and biorobots, as both have their advantages and limitations.