Wet information processing with Maurits de Planque

So often it appears that nature has pipped technology to the post, and information processing is no exception. Scientists are creating a new kind of ’wet’ information processing technology capable of mimicking the neurons in the brain. Here we speak to one of the men behind the new technology.

Scientists at the University of Southampton have received an €1.8 million grant to develop a new kind of information processing technology inspired by chemical processes in living systems. Biochemist Dr Maurits de Planque and Dr Klaus-Peter Zauner, a computer scientist, will adapt brain processes to a ‘wet’ information processing scenario, which copies some key features of the neuronal pathways in the brain. The resulting computer will be ‘wet’ just like our brains and will be capable of excitation, self-repair and self-assembly.

The scientists – from the School of Electronics and Computer Science at the University – received funding from the European Union’s Future and Emerging Technologies Proactive Initiative, which recognises ground-breaking work which has already demonstrated potential. Here we chat to Dr de Planque about the new technology.

What was your inspiration behind developing the 'wet' information processing system?

We all have very different backgrounds, ranging from biochemistry to computer science, but after the EU asked for ideas to develop a 'wet computer', we realised that the droplet system combined all our interests: microscale engineering, synthetic biology, chemical excitation and information processing. But ultimately we got the basic concept from the ultimate wet computer that is already around: our brain.

What is the main objective in this project?

The main objective is to explore whether a chemical 'impulse' that travels through a network of microcompartments (the droplets) could be used for information processing.

You'll be setting up chemicals in a tube which behave like transistors in a computer chip, how will this mimic the chemical processes in living systems?

Our simple system captures a couple of key characteristics from the neuronal networks in our brain. A neuron is a large cell through which a biochemical signal travels from one end to the other, and then the signal has to jump to the next neuron. Our droplets are also individual compartments, and because we fill them with 'excitable' chemicals they can also transmit a signal. We actually stabilise the droplets by coating them with the same biomolecules that coat neurons. The molecules really want to be localised at the interface of the droplets, so if you perturb the system, these lipids will just restore the initial configuration. This self-assembly behaviour also mimics the plasticity and self-healing properties of the neurons in our brain.

The work involves three complementary objectives - could you tell us a bit more about these objectives?

First we need to create the droplet networks and make sure that a chemical signal can travel from one droplet to another. We don't want the signal to jump too easily between the droplets, but it should also not be too difficult. We aim to create branchpoints, where a droplet gets chemical input from more than one droplet, and depending on the exact chemical conditions, will then give a chemical output or not. The next objective is to measure how the real droplet network behaves, and then we will use this data to model whether larger droplets networks, that are closer to the enormous amount of connections in our brain, could be used for computer-like tasks.

How do you see this research being applied in the future?

Conventional 'dry' computer chips are very good at reproducibly and accurately dealing with huge amounts of information. Our brain (or at least my brain) is not so good at this, but we are more flexible: we make non-logical connections that can also turn out to be a great way of dealing with information. We hope that our research will contribute to the design of a more creative computer.