POSITION & HOLD
Tech revolutions and implications for
transportation and society
Where will all these trends and developments lead us?
By Dennis Bushnell,
Chief Scientist, NASA Langley Research Center
Videoconferencing and telecommuting are already commonplace in many areas of endeavor. Like many other technologies, they are likely to become ubiquitous.
It is well known that we are currently in the midst of simultaneous IT, bio, nano, energetics and quantum technological revolutions. The first of these to appear, in the late 1950s, was the IT revolution, followed by bio in the 1970s, nano in the 1990s and energetics and quantum more recently.
All of these technologies are at the frontiers of the small. All are highly synergistic and all are in a "feeding frenzy" with respect to each other. This discussion will attempt to summarize the status and outlook for the effects of IT—the "greatest among equals" of these—on society and transportation.
We have advanced, on silicon, some 8 orders of magnitude in computing since the late 1950s. Going forward, we will go beyond silicon to bio, optical, quantum, nano, molecular and atomic computing, with estimates of some 8–12 orders of magnitude still to go in the next decades.
A petaflop is a measure of a computer's processing speed and represents 1000 trillion floating point operations per second. The nominal human brain speed is some 20 petaflops, and a machine with that speed is to be delivered in 2012. Exoflop machines are in design. [An exoflop represents 1018 floating point operations per second.]
Quantum computing is "different," with initial versions extant and work ongoing to address issues of "decoherence." Estimates of seriously interesting quantum capabilities range from 10–20 years out. For an increasing number of applications, quantum computing is up to some 1044 better.
Quantum computing may enable direct numerical simulation of turbulence in a design mode at realistic flight conditions and further reduce in major ways the use of wind tunnels for vehicle design. The trend toward mod-sim versus physical testing has been evident over the past several decades, with the wind tunnels increasingly providing the modeling, verification and validation for the mod-sim codes as opposed to providing flight-applicable design information directly.
There are many "corrections" for factors such as walls, stings or Reynolds number required to scale wind tunnel data to flight. This scaling and attendant uncertainties are not directly required for ab initio mod-sim design. As MJ Lighthill once remarked to the author, "We build what we can compute," and for too long we were constrained in our designs by nearly linear designs and thinking.
With the evolving mod-sim capabilities we can now seriously consider designs that are "deviant" with respect to conventional air vehicles, including externally-truss-braced wings enabling much greater span, lower sweep, less wing weight, large acreage of laminar flow and synergistic propulsion integration. Overall, the potential for some 70–80% fuel burn reduction exists, courtesy of the IT revolution.
Computer speed is enabling, but the real game changer with respect to the IT impacts going forward is "machine intelligence"—eventually at human level and possibly beyond. There are postulations that human intelligence evolved via "emergence"—in other words, make something complex enough and it "wakes up."
There was an initial effort, lasting from the 1960s to the 1980s, in regard to AI (aka artificial intelligence), based on prescriptive if/then statements, involving massive coding and, overall, found to be not very satisfactory. This approach did result in useful "expert system" software. Since then, machine intelligence has evolved along 2 paths—biomimetics and "learning" or "soft computing."
DOE Jaguar supercomputer at Oak Ridge National Laboratory, the world's fastest computer for open science.
The latter involves, for example, neural nets, fuzzy logic and genetic algorithms and is based on "learning" versus prescription. This has been much more successful than the old AI approaches and is in increasing use in many areas of technology and society. However, to this author's knowledge, there are no clear paths forward to human-level computing via soft computing—at least not yet—although many researchers are pursuing such goals.
The apparent "first to the finish line" for human-level machine intelligence is thought to be biomimetics. There are several efforts in progress, such as the IBM Blue Brain project, in which researchers are nanosectioning the neocortex and other parts of the brain, and replicating them in silicon.
Although the director of the IBM project has indicated that he expects to have a capability approaching human via biomimetics within the next 10 years or so, other projections are another 10 years or more beyond this. In the runup, machine capabilities are advancing in very useful ways and are already included in many application areas. As detailed below, advances in nano and robotic technology, coupled with projected advances in machine intelligence, will have major impacts on human society. The signs are readily apparent today.
The other important IT technology going forward—along with machine speed and intelligence—is immersive presence/virtual reality (VR). Bandwidth and haptic technologies are now sufficiently capable to enable 5-senses VR—haptic taste, touch, smell, sight and sound. (Haptic technology applies tactile sensation to human/ computer interaction.)
Early capabilities along these lines have been demonstrated commercially. They will become far better. In the runup, indications are that many millions of people, using flat screens, are now spending more time in virtual worlds than in the real world. In the future, with neurologics developing as rapidly as it is, we can look forward to direct brain connections to provide inputs to the brain, rather than fooling the senses.
With such superb VR, folks can walk down a virtual tropical beach, smell and feel the sand and the breeze, anywhere they want, anytime they want, be anyone they want, be with anyone they want and do anything they want—without being limited to "reality."
NASA is engaged in efforts to provide virtual space exploration—eg, the Moon, Mars—to everyone 24/7. Initial versions are extant. This comes at some 2–3 orders of magnitude less cost than physical space exploration. It is also safer and will come far sooner. This technology arena includes holographic projection.
IT impacts on society and transportation
The IT revolution is changing society, massively and in real time, via "tele-everything."
Take telecommuting, or telework. Nearly 1/3 of the US workforce currently performs some type of telecommuting/work, and the percentage is increasing rapidly. Traffic in some sections is not growing as rapidly as projected earlier due to telecommuting in lieu of physical travel.
Gasoline use has decreased. This is saving energy and time as well as personal and infrastructure costs—all good developments. More than 25% of workers at NASA Langley Research Center telecommute to some extent, and we have a future study to enable many more to work from home.