By: Richard Morley
Many argue that the best predictors of the future are science-fiction writers. Star Trek, 2002, Jules Verne and the rest seem to have a better handle on the punctuated evolution than professional futurists. Some urban fables that indicate that, in the past, many of us have been wrong in predicting the future. They include:
"Heavier than air flying machines are impossible." - Lord Kelvin, 1895
"I think there is a world market for maybe five computers." - Thomas Watson, 1943
"There is no reason anyone would want a computer in his/her home." - Ken Olsen, 1977
"640K ought to be enough for anybody." - Bill Gates, 1981
We can surmise that primary technological trends of the future's manufacturing are in control and optimization of assets - physical, employee, and management. Emerging control and process business segments, such as semiconductors, are in the dark ages of computer-integrated control.
The automotive and computer industries are enlightened with respect to control. How much control do we need? We can argue this point from the view of the capital investments required. At the moment, most of the legacy process segments believe that five per cent of a new project should be intellectual control such as computers, MES and programmable controllers. We would argue that mature industrial segments need 20 per cent. There is now more silicon or software in automobiles than steel. A computer content of 20 per cent of new projects - as process and enterprises mature - is probably conservative. In the United States, there are more personal computers sold than TV sets. And the overall hidden and obvious monies in the U.S. that go into some form of silicon and software are indeed immense; the industrial segment has to catch up to the consumer segment. Prediction: 20 per cent of any new project will be in silicon and software in some form, imbedded or overt.
In the near term, we can extrapolate these changes with some confidence. We know that Microsoft and Bill Gates will continue to have a major impact on the short direction of technology. His recent decisions to downplay the community agreements on Java, his extensions into Apple Computer, and connectivity to TV, are all harbingers of the future. Microsoft and NT will be, at least in the next five years, the fashionable purchase for enterprises. Entrepreneurs, however, are never put down. We have high hopes that Netscape-style operating systems will come into their own, and that our lack of computer power for virtual-reality makeup language will be solved. Bandwidth is being solved with low-orbit asynchronous satellites and the new 100X modems centered around technologies such as Asynchronous Digital Subscriber Lines (ADSL). These technologies enable, but do not solve, our desire to decrease our time to profit with market agility.
I get excited and fall in love quite easily; one of my love stories is with ADSL or similar technology. These US$1.5M transistor modems are reducing the cost of bandwidth to the local facility over standard, twisted-pair telephone lines with six megabits of bandwidth. This means that every building, enterprise, factory and process system is "already wired" for fiber optics. Of course, we need this bandwidth to allow employees to waste time playing doom. Everything will be run over the network, for example, and the worldwide Web will become the defacto of communication means. Utilizing high bandwidth, modern technology allows hope - in the form of Enterprise Resource Planning (ERP) - to be interconnected into the enterprise. Over the next ten years, we will have a handle on bandwidth, computer power, system integration, and even the ubiquitous problem of software and software complexity. Agents and complexity in emergent systems can help us solve the problem of responsive management as well.
Although Java is considered by many people to be an adolescent technology, the idea of having hardware-independent software packages allowing for agility of configuration and utilization in the application environment is a strong siren call. Java is of no use at the moment, but its promise and the direction that it enables, makes it a winner in the long haul. The best advice I can give in the short term is to put Java and Java compatibility into every element and component of your control system. Deal in systems - not components - and deal with your process, not the love of implementation technology. Our job is to make better products; take advantage of what is, not what you wish you had.
In addition to software, there are other technologies. We can ramble on about autonomous agents, rule-based browsing viruses, self-organizing smart systems, predictive modeling, and real time. And the materials revolution is coming on strong, particularly in the composite and synthetic arena. Linear motors and exotic drives (the muscle that attaches to the brain) are also areas of strong opportunity. Sensing, and the ability to make measurements that inter-relate between systems, is also part of the future.
One of my hobby projects - utilizing Schneider Automation programmable controllers, NT and the Web - is my Case 580C backhoe, which serves as my garden tractor. I expect to have this 22-foot-reach backhoe running from a laptop from anywhere in the world in real time. A sign on the side of machine reads, "Netscape inside." In other words, I'm putting my hobby money where my mouth is. We are dealing with a system. The valleys of low technology can block progress; in my backhoe project, my main problem is the hydraulic components, not the electronics and the Web.
I'm reminded of a trip I made about five years ago to a university in a third-world county, where I was given a demonstration of a system that allowed me to "etch a sketch" of a simple part on a Sun terminal screen. Upon completion of my sketch, I stepped outside the room and they handed me the completed part from the adjoining room with no intervening software or personnel required. Five years later, this is still an impressive demonstration. Stepping off the university grounds, however, I noticed people were dying in the streets from cholera. They could produce a part, but were unable to stop an epidemic. It is not the peaks of technology that we should worry about, but the infrastructure and support that enable these peaks of technology to move forward.
Speaking of holding the universe together, modern systems (in the last two decades or so) and modern software (1960 +), have the characteristic such that, if ants were designed that way, the following would occur. If I step on a single ant, all ants all over the world would cease to move, or at least become jerky in their movements - clearly, a bad scenario. We have the problem of top-down design. We believe that the goals, and not the properties of a system, are more important than the successful and robust execution of system requirements. Horizontal or object oriented technology - or better yet, "bottom up" agent technology - tends to make each ant in our system world loosely dependent. This characteristic leads to extremely robust systems and a substantial reduction in software. Acceptance of new systems technology is the key to agile competitiveness in the future.
I like to make predictions about the year 2020. First, computers will continue to evolve by a factor of 10 every five years. This is known as Moor's Law and has been tracked diligently by pundits since the year 1958. In the year 2020, JAVA will mature. Objects and emergent systems will allow us 4-D databases, and automatic communities emerging from MIT (Massachusetts Institute of Technology), such as Firefly, will become a hot item. Available bandwidth by then will be an established fact with the ADSL technologies, and the low-altitude asynchronous satellites. To leave my conservative limb for a moment, the technology, such as Nan machines and quantum effects, may lead to a Cambrian explosion of technology by the year 2020. In other words, the rate of change is accelerating and non-linear.
What will be the killer applications for the technologies that we project for the year 2020? System integration across the entire enterprise will be an accomplished fact. Right now, we seem to have control over financing, management, and manufacturing. Certainly, the management and integration of engineering into the enterprise is poor. All of those functions will make ERP and MES a direct benefit and not just an application for cosmetic hope. Entertainment - whether games, movies, interactive design, or communications - is a strong investment opportunities for the year 2020. The two classes of security will be well satisfied by then. These two classes are the protection of property and person; and psychological sanctuary. By the year 2020, the applications in financing, stock buying, outsourcing, purchasing, distribution are all going to be on the ubiquitous Web with very high-powered computers, bio-connected with other computers, satisfying instant gratification.
A standing joke about the Internet is that instant gratification is much too slow. Our games will make war more prevalent, quicker and uneven. Forays into space will continue, albeit at a slower pace. While technological changes occur every two to five years, cultural acceptance can take three or four decades, and the gap between cultural acceptance and technology will continue to grow. With these new technologies, and the beginnings of new killer applications, we can shorten the time to profit and increase our competitive agility in the marketplace. These technologies and cultural attitudes will allow us to redesign our products, processes, and relationships with people so that we finally get it right. From this view, the year 2020 is the "get it right" year.
Computers are, of course, the next decade of re-invention, but an understanding of systems and physical science, and its coupling to biology, will become a thrust of the next millennium. Our customers will be served well in the next two decades by computer technology and a maturation of management sciences. Beyond that, who knows? But certainly the future is exciting, interesting, and positive. If we can keep the cultural aspects and its associated frameworks in league, we can maintain a stress-free work place and ride our Harley Davidson - or whatever our pleasure - on the weekends.
|R. Morley Incorporated
614 Nashua Street, Suite 56
Milford, NH 03055-4992 USA
Tel: 603-878-4365 FAX: 603-878-4385