The Future of Research

 Last update: 2005/03/08

There is more than one reason why I feel the need of a page on the future of research. The first is that I have spent all of my professional life, so far - 35 years if I include my Ph. D. in Japan - working in a research environment, and in a period of time where the nature of research has changed profoundly. The second reason is that there is a strict proximity between research and standardisation à la MPEG, actually the two are time-sequential with a long overlap and a strong interaction at the time research results are fed into the MPEG process to be transformed into standards for exploitation. The third reason is that too many CEOs have installed too many brainless bureaucrats at the helms of research establishments, so that industrial research is on the verge of disappearing without anything else taking its place. Lastly, because in the last 15 years I have managed the process of conversion of hundreds of research results into standard components of some industrial interest, I have got some ideas about what can be done to save this fundamental function underpinning the progress of mankind.

It is an obvious observation that the model established by Thomas Edison and perfected by the Bell Labs does not seem to provide value for money today. In this model, research was an undertaking designed to produce innovation leading to new products and services. It worked well with T. Edison because he was the inventor and the rest of the organisation was just at his service. In the Bell Labs case, outstanding inventions were delivered for exploitation by the Bell System that included Western Electric, the Bell System's manufacturing arm, in addition to ATT. The model used to work well because at that time the task was to find technical solutions to problems that corresponded to new technologies enabling new products, like electronic tube and loudspeaker, to process improvement, like the the transistor replacing the bulky electronic tubes or UNIX, a generic OS replacing the plethora of different OSs or to well-identified problems like adding sound in motion pictures or mobile telephony.

The Bell Labs model was copied by other telcos in Europe and Japan, however, the research establishments of the other telcos did not, in general, produce anything comparable to the string of inventions of the Bell Labs. But these establishments did provide the technical means to plan and design the evolution of their networks from analogue to digital - again a well-identified technical problem - with the added advantage that the Bell Labs was the clear leader and all the others could be the followers, whether they were ready to declare it or not. The process ran smoothly because funding of research happened as a sort of internal tax, such as 1% of the company turnover.

Public broadcasters in Europe and Japan adopted a similar model. The budgets assigned to broadcasting research were in general much lower than those assigned to telco research, but the field to be covered was also substantially narrower. Besides, the headquarters of broadcasting companies tended to be more conservative in using innovation because every change has to be assessed and endorsed by appropriate political circles. Only a rather small number of selected areas were attempted: colour, teletext, MAC, HDTV, MUSE, HD-MAC, digital modulation, ISDB, not all of which were successful and not so much because the technology was not good, but because of the litigious nature of that industry.

Research played a major role at IBM, the model company and the undisputed market leader in the first decades of the computer age by providing the means to sustain and extend the leadership acquired. But it was again the case of research being asked to provide solutions improving processes in response to precise market needs, because the success rate of more speculative innovations is not equally exciting. It was hard for other IT companies to apply the IBM model  because of the less stable flow of revenues that smaller players enjoyed. 

At the beginning, the role that IBM played in the IT world was played by RCA and Philips in the CE world, even though the two never achieved IBM's imposing market leadership. Also in this case, research was asked to provide exploitable results in a sustained fashion to keep consumers' interest in new gadgets awake. In the early days this was not so difficult if one thinks that consumers lived in houses that basically lacked the sort of electrical and electronic appliances that populate our homes today. With the decisions, first by Japan and then by Korea, to become major players in the CE market, the Philips model was thoroughly implemented by Japanese and later Korean CE companies as well.

The situation today has radically changed. Setting aside the undisputed success in providing the fixed and mobile telephone service to ever wider swathes of population, the telco industry has basically missed all other major targets: ISDN, OSI, ATM, videotelephony and broadband services. This does not mean that no innovation was introduced in the telco business, but if anything, this happened largely in the infrastructure. Operation and management of the huge telecommunication system has improved over the decades, but what major innovation generated by a telco's research establishment has become a traceable advantage practically exploitable by the telcos? Even such "internal" industry projects like the Telecommunication Information Networking Architecture (TINA) failed. 

The disconsolate conclusion is that telcos understand what their network needs, they do not understand what their customers want. The result is that, at the service level, they simply go after the perceived whims of their subscribers, biased by such deep-rooted and naïve misconceptions as visual telephony for the masses. At the infrastructure level they keep on squeezing the system by optimising its internal workings. Unfortunately, there are physical limits to what you can squeeze out of a system: sooner or later if you want to increase the output you have to increase the input, i.e. provide meaningful services that people are willing to pay for.

The success rate of broadcasters' research establishments is possibly even lower. The record of successful exploitation of CE research, setting aside very few but significant cases, is low. IBM's research is still important because research generates IP that is not just acquired by IBM's business units but is also licensed outside. Research carried out by other major players in the IT space is more a means to find out what to do next when the current franchise will run out than a support to sustain the current business.

It can be understood that CEOs feel frustrated by the inability of their companies' research to provide results that give value for money, but their actions cannot be. If a research establishment does not provide the exploitable results they expect, the solution is not to appoint an accountant with the mandate to cut budgets as the head of the establishment, but to devise a process by means of which research provides benefits to the company. I know that the recipe is obvious, but cutting travel budgets is meaningless. It would be a show of courage, if not of capability to put excellent brains to good use, to close research altogether.

Who else should be appointed then? The obvious remark is that it should be somebody who knows both what products the research establishment can offer and also what products its customers need. Unfortunately this is not the usual customer-supplier relationship, because I still have to see a thriving private research establishment that is able to live on the revenues of research results sold on the open market. The solution is, as for all matters that involve human beings, to create a virtuous circle whereby innovation is rewarded because it is adopted by a business unit, adoption of innovation is rewarded because it brings measurable benefits in the operation, efficiency or rentability of the business unit, and innovation is generated because there are people in the business units who are smart enough to fund the development of innovation because they know their smart choices will be rewarded.

I am afraid nothing of what I advocate corresponds to my experience. So CEO-appointed accountants-turned-heads-of-research keep on issuing orders to cut travel budgets by another 5%, thinking that with this, research will finally be channeled to business units.

In past centuries bloodletting was the prescription for most illnesses. In those primitive times not much was known about the human body and patients were often in the hands of barbers for whom bloodletting was the principal cure, as it was based on their enabling technology - the razor. Today a doctor prescribing bloodletting would be regarded with high suspicion because we know we can do better than that. In the same way it is high time that management in companies stop using cost cutting in research as a means to hide the vacuum in their brains.

If you do not know what to do with research don't cut travel expenses by another 5%. Take responsibility and close down the establishment. Then go to the Board of Directors to explain and hope I have not been appointed a Board Director in the mean time.

Copyright © 2003 chiariglione.org