Carnivorous robots and digital plasters
January 23rd, 2008
Last night I attended the Dana Centre to hear a collection of scientists discuss the blurring boundaries between technology and your body. Contributors from the Bristol Robotics Laboratory, the Future of Humanity Institute at Oxford University, the Institute of Biomedical Engineering at Imperial College London and University of Reading’s Cybernetics group presented the latest research from their respective departments.
Of the four groups, I was most interested in EcoBot, a project to create an autonomous machine that would take energy from its environment. Unlike solar- and wind-powered devices, which are dependent upon the right environmental conditions, EcoBot would hunt down its own food, advancing robots on from imitating plants to imitating animals. EcoBot-I was designed to hunt slugs, because, in Prof Alan Winfield’s words “Firstly, nobody likes them, and secondly, they’re very slow”. The finished was almost primordial in design, with a long thin neck, large jaws, and a stout body (video). This allowed it to hunt the ground for slug without having to move very often, saving energy. Contained within the tri-jaw was a camera and a red light that would illuminate the slugs (which do not show up on infrared, as they take on the temperature of their environment). Having proved a robot could predate efficiently, the next step was a robot that could digest the captured food.
This is where it gets exciting. The black boxes arranged in a ring are microbial fuel cells – in essence, small stomachs that can digest organic matter and produce electricity. Arranged in series, each generates a few microwatts of power, enough to fuel a simple brain and light-seeking behaviour in EcoBot-II. The best food source for EcoBot-II turned out to be chitin – the principle protein polysaccharide found in the exoseletons of insects. In fact, a single fly in each stomach is enough to power the robot for two weeks! Unfortunately, the waste products of digestion eventually kill the essential bacteria in the fuel cells, rendering them useless.
Currently in the pipeline is EcoBot-III, which uses a trap to catch flies (much like the Pitcher plant). These are digested and the juices flow through the microbial fuel cells and drain out, keeping the levels of toxins at a tolerable level. What it will use the generated power for isn’t yet clear.
Prof Winfield imagined these robots as roaming predators, eating pest species in your garden. My question to him was: why bother when there are already far more efficient biological pest control species in use (e.g. the use of ladybirds to combat greenfly)? Wouldn’t it make more sense to design a robot that filled a gap in the ecosystem, perhaps by designing it to digest oil, plastic, or toxic or radioactive waste? His answer (if I can paraphrase correctly) was that these machines represented an industrialisation of natural processes, just as planting corn in rows improves the efficiency of farming, so too could these robots be more efficient than natural pest control measures. We didn’t have time to discuss it much further, but I really want to know the answer to this next question: if we can create robots that are autonomous and biologically embedded in the ecosystem, what are the implications of releasing these hybrids “into the wild”? The introduction of non-native species can play havok on an ecosystem – could EcoBot be the next grey squirrel or Japanese knotweed?
Further reading – you can visit Professor Winfield’s eloquent blog on all things robotic here!
Part II – digital plasters – will be uploaded soonish. Are tiny sensors that monitor healing and transmit vital stats a good thing? Or is it another step in making private information available to anyone with a scanner?
Entry Filed under: General
|
|
|
|
9 Comments
1. badchemist | January 24th, 2008 at 12:05 am
Dear me Frank, chitin is a polysaccharide not a protein. Very cool robots though.
2. Frank the SciencePunk | January 24th, 2008 at 12:45 am
D’oh! Schoolboy error! (I blame a failure of chitin to follow typical nomenclature for polysaccharides – glucose, sucrose, etc., and instead take on the protein naming convention – albumin, insulin, keratin…)
3. Science_fox | January 24th, 2008 at 11:31 am
I’d love a slug eating robot for my garden. They may have natural preditors but they are not doing a very good job where I am.
I’m not sure that releasing robots into the wild would be a good idea, but for small localised areas – a garden / factory etc they would be better than importing bio-solutions because you should be able to prevent them spreading/reproducing. Something that biology is remarkably good at.
4. manigen | January 24th, 2008 at 1:41 pm
“you should be able to prevent them spreading/reproducing”
I think it’s the reproducing that’d be the key difference between a robot and japanese knotweed. After all, for one of these robots to make another one they’d need to be able to smelt.
5. Alan Winfield | January 24th, 2008 at 10:28 pm
Glad you like the Ecobot.
Re your question: if we can create robots that are autonomous and biologically embedded in the ecosystem, what are the implications of releasing these hybrids “into the wild”?
If we could make versions of these robots that could ’survive’ in the wild, they would still have two fundamental limitations. Firstly, they can’t reproduce, and secondly they cannot repair (heal) themselves. Furthermore, they would not be very smart – they would just have a small number of instinctive behaviours and wouldn’t be able to learn or reason. Thus, as soon as any of their components failed, the robots would simply stop working and – as it were – die. There are other reasons this might happen – the robot might get physically stuck, or damaged by a real animal, for example. Ideally the robots would be built from bio-degradable materials so that when they did stop working they would – like real animals – just rot away.
There is perhaps another aspect to consider, which is that Ecobots might be competing with real animals for the same resources. In that case, and if there were enough Ecobots, then I guess it might be just possible that the natural species could be threatened by the Ecobots. I would bet however, that the limited population of Ecobots (that can’t reproduce) simply wouldn’t be efficient enough predators to make much difference to the overall slug, or fly population.
Of course if we could make self-replicating robots (sometimes called Von Neumann machines) – which is a profoundly hard thing to do – releasing these could have potentially extremely serious consequences. Check out the wikipedia entry on self-replicating machines for a very good review.
6. Frank the SciencePunk | January 25th, 2008 at 10:49 am
And of course, where there’s competition, there’s evolution, so it’s not impossible to foresee the animal population evolving in response to added pressure from Ecobots.
In fact, it raises an interesting idea in terms of animal husbandry: instead of selective breeding, we might make machines that would compete with animals in ways that would encourage them to evolve toward what we wanted. Hmm…
7. Sarah Summer | January 31st, 2008 at 10:38 pm
My friend would build robots that ran on bacteria while he was at school–the more deadly the bacteria the better electricity it made.
8. Frank the SciencePunk | January 31st, 2008 at 10:51 pm
Pictures or it didn’t happen.
9. Sarah Summer | July 21st, 2008 at 5:29 am
Thanks for this info! I love robots!
Trackback this post