‘The way to do research is to attack the facts at the point of greatest astonishment’ Celia Green 1915    

In this issue is a review of the methods used in the mining and metallurgical industry to manage the research portfolios of a crosssection of SA companies. It makes interesting reading, but I must admit that I battled to comprehend the significance of some of the typical business school diagrammatic representations of risk in portfolios...  There are many ways of killing a cat and of course the companies must elect according to their own culture what methodology they establish for their R&D expenditure. If we take out the ‘D’ from R&D (which relates predominantly to adaptation of existing technology to production and marketing improvements), and focus on the real innovative Research, then some comments on a strategic national portfolio of research might be appropriate.

The first comment, which is fundamental, is that I question whether there is any scientist in the world who can predict where the advancement of technology is going to lead us beyond a few years or assign a high probability of success to a truly innovative programme. There are many who believe differently and at one time it was fashionable to undertake a 5-year advance forecast of what technology would be topical and profitable so as to decide where to put one’s money for research. Most of the first world hitech countries did this on a national basis and we in SA did the same.

I have yet to see any of these forecasts put to the test for accuracy after the 5 years. There are many scrapped remnants of facilities around the world of huge expenditure on innovative research concepts that have been abandoned before the targets were reached. Perhaps the most opulent example, which was enthusiastically undertaken by the top first world technical powerhouses using the most renowned scientists and engineers for many decades and promised huge benefits for mankind, was the attempt to generate thermonuclear power. The basic requirement seemed simple. It was to raise the temperature of tritium to the same levels as in the sun when a nuclear fusion reaction would occur, releasing vast amounts of energy.

It was confidently predicted that after the successes achieved with the uranium, plutonium and hydrogen bombs, it was only a question of time and money to achieve this ultimate answer to the world’s energy needs. But the universe and nature on this planet are unpredictable, and success in innovative research is forever only a statistical probability. We in SA, together with all other technically capable nations, have our failures and the reasons for failure are many and varied but not uncommonly related to underestimation of cost factors and unexpected changes in market demands. The point I wish to make is that there is no shame in accepting that innovative research is a statistical matter and for every success there can be many failures. Unfortunately the business and political executives usually believe that by using superior management techniques they can regularly select winners, and failure implies technical incompetence on the part of their researchers.

A much better approach is one in which the successful project can pay for a number of aborted ones and the art of research management is to provide a portfolio in which the cost/benefit statistics are in one’s favour. I will confess that, in my 56 years of association with longrange innovative research, roughly only one in ten major projects has produced the big payback. But this ratio varies around the world and in different industries. In the drugs and pharmaceuticals I have heard of figures of one commercial hit in 100 starters. I must emphasize that I am talking of innovative research, not blue sky fundamental work, but applied research aimed at producing benefits to the investor, albeit in the longer term. From concept to commercialization often takes well over 10 years.

Development work is in a different category and can and usually does have a much higher success level. The successful R&D manager must ensure that the long-term pipe dreams do not throttle the payback from development work. To do this on a quantitative forecast of costs and benefits requires more perspiration than inspiration. The skill is to identify the high risk, make or break components of a project and to tackle them first so that the project can be aborted at an early stage if these critical steps cannot be achieved. This is analogous to the selection of a share portfolio in an investment company where invariably it is recommended that one spreads the risk over a range of shares from high risk to low risk on a roundabouts and swings basis. This is relatively easy to do on the basis of dividend yields, earnings per share and beta values.

Although the uncertainties are much greater in R&D projects, a similar analysis can be done using rigorous conceptual analyses with definition of critical experiments, critical path planning and probability based cost/benefit evaluation. This can be computerized for regular updating and review. The vital feature of this discipline is that, like a share portfolio, one can get rid of a non-performer and replace it with a new selection or expand activities on a more promising option. If one works through the mathematics this regular revision very much improves the overall portfolio probability of success and a predicted ROI for the whole portfolio. Another important consideration is that the ‘Research and Evaluation Team’ must comprise far more than a group of creative research boffins.

It must comprise hard-headed pragmatists who can distinguish competence from wishful thinking and experience in feasibility and cost evaluations. One needs contact specialists with customers and markets and competitors. There are not many commercial organizations in South Africa who can afford the luxury of a multi-component R&D portfolio as I have suggested, not even in our once mighty mining industry. This probably explains the prevalence of the ‘focus’ approach of selecting one or two potential winners and concentrating all effort on them.

The inevitable statistical failures explain why there are so few innovative research laboratories in private industry. In a research backwater there is an alternative approach, which is to promote government-based research organizations that have the staff and facilities to maintain a worthwhile portfolio of long-range research projects. This also implies support of university groups. In a previous Journal Comment, I suggested an approach in which industry can become stakeholders in such a portfolio and in this way we can become a global player in scientific research activities. I believe that it is important that we seriously move in this direction. We have more than a few globally important resources where there are exciting research portfolio challenges and where we are only scratching the surface.

These are enviously regarded by many others, not only to ensure a supply of basic commodities at low added value prices but also to maintain a high level of further processing capabilities for their own advanced employment opportunities. This is perhaps just another form of global colonization, not politically nor geographically but in terms of technology sophistication. Although small in terms of research capability, we must work cooperatively with all our facilities, private and government, to match this sophistication. I often wonder whether the preoccupation with developing individual ‘scientific entrepreneurs’ should not be second to enhancing excellence in multi-specialist teams allocated to national portfolio activities.
R.E. Robinson  October 2006