SABRE - Self-healing Cellular Architectures for Biologically-inspired Highly Reliable Electronic Systems

Objectives

Powerful and sophisticated systems, from computers, through control systems, to 'conventional' household appliances have become a necessity in our modern way of life. In the modern world of digital electronics - virtually all of which is now built using VLSI technology - we are quite accustomed to the idea that hundreds of thousands, often millions, of individual components on a chip must work faultlessly over extended periods of time. Yet, it commonly requires only a single transistor to fail to have catastrophic consequences for the entire system.

Imagine an automobile travelling at high speed in the fast line of a busy motorway. Suddenly, the electronic Engine Management Unit (EMU) develops a malfunction, the engine cuts out; soon after this the servo-assisted brakes and steering (that depend on the engine inlet manifold vacuum) both cease to function properly; a queue of stationary vehicles is fast approaching. The outcome of this scenario is left to the reader's imagination. This is a somewhat dramatic thought experiment, but one that brings the safety-criticality and reliability aspects of some everyday digital electronic components in our lives into stark focus.

The design of complex, but reliable, electronic systems and ensuring their long-term fault free operation is a major challenge we face today. This demand is even more pronounced in the case of electronic systems where their correct operation is imperative, e.g., anti-lock braking systems, fly-by-wire aircraft, space exploration, industrial control and shutdown systems; they should be able to operate correctly in the presence of faults and be fault tolerant.

How can we design such reliable systems? Nature offers some remarkable examples dealing with complexity and unreliability. Living organisms, and in particular the human body, is one of the most complex systems ever known. Yet they possess an extremely high degree of reliability. Although local failures, due to harmful pathogens and environmental conditions, are common, the overall function of the organism is highly reliable. Many of the cells and tissues die as a result of damage, but because self-diagnostic and self-healing continues incessantly, full functional integrity of the body is not compromised. It will carry on working properly because the body's defence mechanism, comprising numerous immune responses, will try to restore its full functionality. We could therefore justly ask ourselves the question; "would it be more efficient and less costly to draw inspiration from nature in how it deals with the complexity vs. unreliability issue with such a remarkable degree of efficiency?".

The challenge we propose to take on, therefore, is to adapt biological processes found in living beings in our pursuit of designing reliable electronic systems that demand an ever increasing level of complexity. Although a great deal has already been achieved in these areas, much of this progress having been made by the two collaborators in this proposal, there remains still a vast amount to be done.

Summary

The aim of this project is to further our understanding of how the multi-cellular nature of biological systems and their protective immune systems could be used to enhance reliability of digital electronic systems, focusing especially on innate and acquired immune responses, and on incorporating these 'nature/like' fault-tolerance mechanisms into engineering systems.

We will fulfil this aim by achieving the following objectives:

Produce efficiently implemented fundamental cells with intrinsic self-repairing and growth mechanisms;
Use multi-cellular development to create a new design approach based on cooperating cell groups;
Use the combination of an innate and acquired immune system to supervise the self-repairing architecture at both intra-cellular and extra-cellular levels, so as to identify, predict and protect against errors;
Use FPGA (Field Programmable Gate Array) based implementation to construct a demonstrator to illustrate the capabilities of the system in a service-robot application scenario.

The above tasks will be approached from two different, but convergent, levels. At the cellular level, we will consider the basic building blocks of our system and its intra-cellular repair mechanisms. At the organism level, we will consider integrated multi-cellular structures that together perform some given application level functionality and provide repair mechanisms that operate across the whole organism. Between the two there is an intermediate tissue level formed by clusters of cells with autonomous tasks that will be the point of convergence of the collaborating partners' work. This third layer eases integration of the cellular and organism structures, allowing us to consider extra-cellular structures that span across a number of cells but not the whole organism.