Computer-Aided Innovation For ICT

Background

The ICT sector has seen unprecedented growth since the invention of the semi-conductor chip. Furthermore the rate of growth still seems to be on an exponentially upward trajectory. The consequences of this growth include ever-shortening product development timescale needs and ever more rapid obsolescence of existing technologies. It thus becomes progressively more and more difficult for companies to build and maintain a lead over competitors. The problem is only exacerbated by the growing threats provided by India, China and other developing nations. These are nations with the twin benefits of an enormous drive to generate the revenues that will pay for the nation’s development, plus a highly educated population with a significantly lower cost structure than can be found in the UK.

There are many other consequences of rapid ICT growth. The one which acts as the focus for this project is the increasingly daunting task of managing knowledge in such a way as to create the right context for future innovation. A first part of this task involves identifying and coding existing knowledge into a framework that allows both ready access and the ability to rapidly transfer good ideas from one sector to another. A second then involves establishing when and where a given existing piece of knowledge is no longer relevant to the prevailing market and organisational context.
The platform for the research is an existing Systematic Innovation methodology. This methodology has thus far accrued data from over three million innovation case studies from all forms of human endeavour. The main findings of the research are:

- Major innovations occur when ideas are transferred from one industry to another
- Major innovations occur when trade-offs and compromises are resolved
- Major innovations all follow distinct patterns of discontinuity that repeat across different industries

As a result of building this three-million strong repository of innovation knowledge, IFR Consultants Ltd has been able to build an impressive roster of global clients. Amongst the companies we work with are Intel, Procter & Gamble, Samsung, Rolls-Royce, Motorola, GE, Infosys, Hilti, SRA, Telekom Malaysia, General Motors and, through the Hong Kong government, a large number of Chinese SME companies.

The main reason these clients use our services is that we offer them a bridge to the new ideas and successes in other industries and the ability to therefore accelerate the innovation in their own industry. A key success factor, therefore, is that we are able to demonstrate that we are continuing to track and incorporate new knowledge as it emerges. At the present time, this updating is done through a dedicated team of 20 researchers. This team focuses primarily on the patent databases of the world, and their job is to distil useful innovation knowledge from the strongest IP. An emerging problem in conducting this activity is that things are moving so quickly in and around the ICT sector that new domain knowledge is required in order to be able to understand what is useful and what is not as new patents emerge. In particular we have identified three areas where there is a need for a specific focus: 1) nano-technology (and in particular in the area of semi-conductor design and manufacture and hardware implications), 2) complex-systems (specifically the need for understanding of the dynamics of massively-connected networks), 3) the progressive evolution of systems away from the mechanical and towards ‘field’-based design solutions (magnetic, laser, nuclear, etc). The aim of this project, then, is to build a systematic knowledge distillation and modelling capability in and across these three domains. Once these new technologies can be adequately mapped it will become possible to extend and refine the existing three-million strong knowledge repository and thus be able to provide organisations in and around the South West with access to a unique new knowledge creation capability.

Intimately coupled to this knowledge acquisition objective is the need to provide organisations with ways and means to at least maintain the innovation rates being achieved by overseas ICT players, but preferably to provide the capability to drive the technology advance. The ability to drive rather than follow is known to be important economically (John Kao, ‘Innovation Nation’, October 2007). It is also known that one of the critical barriers to success is finding and protecting the necessary manpower resource. The West in general and the UK in particular faces an enormous and increasing challenge in this area, since we can only hope to acquire a tiny fraction of the resources becoming available in India and China – both of which are currently graduating 4 times the total number of ICT professionals than exist in the whole of the UK every year. The answer to this growing manpower-innovation gap is increasing automation of the innovation process. ‘Computers that invent’ in other words.

The latest generations of computers are already fully able to conduct design optimisation tasks. This is an important capability in a ‘continuous improvement’ lead economy. It is not sufficient, however, when, as has happened in ICT, the paradigm shifts to one of ‘continuous innovation’. Computer-Aided Innovation (CAI) is a subject that is still at the most formative of stages. To date there have been just two academic conferences around the world convened to discuss the subject. We have been present at and presented at both of these events, and, thanks to our three-million data-point knowledge repository, appear to be uniquely positioned to turn the theory into some kind of practical reality. A further aim of the project, therefore, is, once the new nano, complex and field knowledge has been added to the knowledge repository, to begin to assemble and test computer-assisted innovation tools and strategies. What we already know about the subject is that the biggest challenge, once a knowledge repository exists, is to allow the computer to understand context. Computers are already able to ‘invent’ – i.e. make design jumps (e.g. genetic algorithms) – but they have no way of knowing what is a useful jump and what is a bad one. This context-mapping is something that currently only humans can do. Computers are able to make millions of random jumps, and so very quickly the new human problem becomes having sufficient time to sort through the jumps. What we now know from our three-million data-points is that of the millions of random possible jumps, there are only actually a very small number (around 40 at latest count) that will result in meaningful, context-relevant outcomes. With this in mind, one of the things we expect to deliver from the project are prototype computer-assisted innovation tools based on these 40 successful jumps.

As a final stage of the project, we will be working with world-class industry partners to conduct a number of CAI validation case study assessments. The aim of these case studies will be to verify the effectiveness of the tools and strategies defined in the project, and to moreover, deliver useful, protectable and economically viable design innovations to the partner companies.

You
The details of the research programme have been left deliberately open in order to give the chosen researcher the opportunity to build their own thoughts and concepts to the programme. As such, the success of the programme rests on an individual capable of thinking about big pictures as well as fine details. A wide number of different backgrounds may form a suitable platform upon which the project will build. However, it is likely that the successful candidate will have done some work in the artificial intelligence world, and will have a working knowledge of NetLogo or equivalent and VBNet.