Swarm Engineering: From Flocking Robots to Nanomedicine

How can studying swarms lead to more efficient buildings and better medicines?

I was lucky enough to attend a lecture by Dr Edmund Hunt at the Rutherford Appleton Laboratory near Harwell last week. Dr Hunt took us on a journey of bio-inspiration to show how the study of nature may lead to the production of novel nanomedicines or self-organising robots that become capable of solving complex problems.

Many animal species produce swarms. Bees, fish, birds and ants, for example, all show social behaviours which allow the group to perform tasks that individuals simply wouldn’t be able to do. Insects often work together to build huge nests and then ensure their success by looking after the egg-laying queen and her young.  Birds and fish will often move around in vast coordinated groups, which have the effect of confusing predators and creating a degree of safety in numbers.

By studying some of the simple rules that each individual follows when acting as part of a swarm, scientists can program robots or create animated video sequences which behave in similar ways. It turns out that we can model swarming behaviour by using just 3 such rules shown by starlings in flight. These Reynolds Flocking rules can be summarised by the diagram here.

The rules are straightforward. Each bird needs only pay attention to its immediate neighbours and can ignore the behaviour of the rest of the flock. By introducing a few other rules such as “avoid flying into obstacles” we can describe flock behaviour fairly accurately.

The stampeding wildebeest scene in Disney’s Lion King was animated using these rules to produce incredibly lifelike sequences. Watch how the individual animated creatures stay close and align themselves to each other while running around the various rocks in their path.

Why do these animals act as swarms in the first place?

A group of individuals acting as a swarm can do so much more than any of the individuals can do on their own. To use a phrase from Aristotle: The whole is greater than the sum of its parts. Alternatively, any player or coach of team sports will recognise the acronym TEAM as meaning Together, Everyone Achieves More. Swarms of individuals can respond quickly to changes in the environment. They are amazingly robust, in that if a number of individuals die, the system as a whole still works perfectly well. The other big advantage is that swarms are very scaleable. It is very difficult to control the behaviour of all members of a group from the top down, particularly when the group becomes reasonably large – just ask any CEO of a large organisation. However, when all the members of a group are relying on the behaviour of its immediate neighbours, it is much easier to effectively organise the behaviour of the whole group – and the group size can grow to huge levels without adding to the problem of controlling it.

So now that we have studied swarming behaviour in nature, what can we do with it?

Development of nanomedicines

Nanomedicines are made from really, really small molecules. By small, we mean around 5 billionths of a metre in diameter, or 5 nanometres. Particles of this size have some unusual and unpredictable effects on living tissues. By treating these small molecules as members of a swarm and varying the set of rules by which they behave, researchers can study the emergent properties of the molecules as they interact with each other and with living cells, and in turn attempt to create a molecule which acts in such a way to be medically beneficial. Such molecules might be capable of killing cancer cells while leaving non-cancerous cells alone, for example. Or maybe molecules could be produced that allow certain chemicals to cross cell membranes to cure diseases such a Cystic Fibrosis.

Nanodoc.org are using crowdsourcing to find out the optimal configurations for some of their potential nanoparticle cancer drugs. They have produced simulation games, in which thousands of public participants can design virtual nanoparticles. These can then be run through the simulator thousands of times to create huge quantities of data which are used to show which nanoparticles have the best chance of treating patients. An animated description of their work can be seen here.

Development of Swarm robotics

By studying how organisms in natural swarms make decisions about their environment, we could build robots that work together in a swarm to efficiently solve some fairly complex problems.

For example, the Biorobotics Institute have a project called Collective Cognitive Robots (CoCoRo) which aims to produce a swarm of self organising underwater robots capable of monitoring and researching deep underwater habitats.

Other potential projects could involve looking at how the simple rules followed by termites lead to the way they collectively produce their large mounds with elaborate temperature control and ventilation systems. Architect Mike Pearce became inspired by termite colonies when planning an efficient heating and ventilation system in a new office/shopping building in Harare. Listen to his story of how studying termite mounds allowed him to produce a ventilation and heating system with significantly lower energy consumption than other buildings.

We might also one day be able to produce a swarm of robots capable of building stable structures which are precisely tailored to the conditions of the surrounding environment from scratch.

Alternatively, we may be able to better understand how a bundle of cells in an early embryo are able to arrange themselves into the various cell types which eventually lead to the formation of a fully functioning human baby.

Swarm technologies are very much in the early stages of development but do offer some exciting prospects. If scientists can find ways to ensure the safety and security of such projects as well as creating user friendly interfaces between human and swarm then the future applications should be limited only by our imaginations.

Current school-aged students are ideally placed to get involved in this new methodology. Teachers are often saying that we are trying to prepare students for jobs that don’t yet exist. Swarmologist might be one of them.

Who knows? Tomorrow’s science may actually take the world by “swarm”.

Pu.

Reynolds, C, (2001) Boids [Online]. [Viewed 19/06/19] www.red3d.com/cwr/boids/