When an asteroid strikes a planet, things tend to change quickly. Most other changes occur gradually, even while a small handful are occasionally more significant. These latter changes are sometimes referred to as "paradigm shifts," largely because they affect so many things around them, and cause such dramatic changes in the things they affect directly.
Driverless vehicles (or "highly-automated vehicles," as USDOT prefers to call them) are unusual in these terms. This is because a completely driverless vehicle is indeed a paradigm shift. (See installment #1 of this series: "Autonomous & Inevitable, Part 1: What is to Come, What is Already Here," in National Bus Trader, October, 2016.). At the same time, there have actually been decades of movement toward such vehicles. Among the most well-known of these activities was the Federal Highway Administration's "Intelligent Vehicle Highway Systems" (IVHS) program. This program actually had its roots in the General Motors Futurama exhibit at the 1939 World's Fair -- initially as a mathematical algorithm. One of the least well-known recent incarnations has been Google's seven-year-old development of this technology hidden away, effectively as a "skunk works" project, in the GoogleX "moonshot" division. Using GPS technology, the project evolved to comprise a method for tracking the speed and direction of vehicle movement, in real time, over a road network.
These projects turned out to be savvy long-term investments. Some day they may be considered milestones. Along the way, decades of technologies have paved the way for a vehicle to eventually drive itself."
Building Blocks and Hierarchies
Summarizing decades of ingenuity, quite a few technologies were integrated into what eventually became HAVs quite a while ago. In their original forms, they were all designed and intended to assist humanoid drivers. Among them, with their approximate years of emergence, were:
- Antilock braking systems (1978)
- Traction Control Systems (1986)
- Parking Aids (1993)
- Electronic Stability Control (1996)
- Adaptive Cruise Control (2000)
- Parking Assist (2008)
- Predictive Emergency Braking Systems (2010)
- Lane Keeping Support (2012)
- Traffic Jam Assist (2015)
- Remote park Assist (2015)
Many other stepping-stone technologies, such as lane-drift technology and retinal detectors (both of which identified possible driver fatigue) are excluded from this list because, without drivers, HAVs experience no fatigue. So these technologies (far more developed than they were installed) were stepping stones soon to be rendered obsolete. Similarly, less-automatic safety-enhancing technologies -- like windshield-mounted video-cameras (Drive-Cam and Safety-Cam were among the first) and parabolic mirrors (Rosco, Tiger, Mirrorlite) -- are omitted because they were designed solely to assist drivers, even though these cameras and mirrors have morphed into sensors within an HAV's vision or guidance system.
One could also argue that other technologies (such as VORAD's motion-sensing technology) belong on this list because their concepts are valuable to an HAV, even while the VORAD system (like most of the technologies listed above) was originally designed and intended to assist drivers. Interestingly, many seemingly age-old technologies (horns, lights, sirens and other bells and whistles) will obviously be present in HAVs in some form -- even while they were never intended for the "ears" or receptors of HAV controllers. Of course, the speed of their activation will change exponentially.
Further acknowledging that the leap from humanoid-driven vehicles to robots did not occur in a single bound, the Society of Automotive Engineers International (formerly SAE) published, in 2014, a classification system reflecting degrees of vehicle automation -- a classification which NHTSA almost immediately adopted to replace its similar 2013-derived classification scheme. This progression resulted in a classification of six levels of automation:
- Level 0: Automated systems with no vehicle control, but including warnings to drivers.
- Level 1: Automated trip segments, but where drivers are posed to assume vehicle control at any moment. (These automated segments include cruise control, and parking assistance with automated steering and lane continuity monitoring and correction.)
- Level 2: Automated features which require drivers to respond where appropriate. Typically, these features govern acceleration, speed, deceleration, braking and steering.
- Level 3: In selected operating environments, drivers can refrain from engaging in many driving tasks altogether.
- Level 4: The vehicle's operation is fully-automated in most operating environments (exceptions include inclement weather), whereby driver attention is not otherwise required.
- Level 5: Other than starting the system and selecting the destination(s) (and in some instances, selecting specific trip segments), there is no driver involvement.
The reader should note that many Level 1 features have increasingly been offered in personal occupancy vehicles for some time. (Level 0 features are often referred to as "idiot buttons.") Plus, some forerunners of Level 2 technology (e.g., "governors") which impose limits on performance or performance failure have been around for decades. Similarly, hybrids of semi-automation have existed for more than half a century -- for example, the supplemental shotgun-seat-located steering wheels and brake pedals installed in vehicles deployed in "driver education" programs.
The central themes of most levels of automation involved the delegation of selected operating tasks to machines, with the ability of a driver to resume control under selected circumstances. The segmentation of these assignments has been much easier for commercial aircraft where, for example, little pilot intervention is required other than during take-offs and landings. Yet even in aircraft, contingencies like inclement weather require pilot intervention in trip segments that are normally automated. Interestingly, tasks that require no human involvement in ground transportation (e.g., idling) are extremely complex in marine operations, where the "roadway" is in constant motion, and exerts tremendous forces on the craft. Holding still is also challenging in aircraft (only helicopters and hybrids like the Osprey can hover). Because of forces like gravity, and primitive inventions like brakes, holding still is only effortless in ground vehicles.
Such principles are obvious, scientifically. Yet their details and nuances are challenging in the rapidly-evolving world of ground vehicle automation, since both the functions automated and their sequence of development are significantly different.
Oops, Back to Start
Those commenting on modern technologies are often reluctant to acknowledge the significance of their often-advanced forerunners. The most salient example is our institutional failure to acknowledge that anti-derailment technology (e.g., flag men and dead man's switches) is practically as old as railway technology itself. A more useful analogy to HAVs is the fact that many of the first pupil transportation vehicles had no drivers. They were simply buckboards which transported a handful of rural students from their families' farms to their schools -- and back again after the horse took a standing nap or two. The wagon master only held the reins on the first day of the school year. Beyond that, the horse pretty much memorized the route. With occasional prodding from a passenger or two, it ran the same route (unless or until it was modified by human intervention) for the rest of the school year without anyone at the reins. (In fact, the bench seating still found in schoolbuses has its roots in buckboard seats.) To not confuse the horse, adjustments were generally not made when a child skipped school and the stop was unneeded (about which most fellow-passengers knew nothing either).
In contrast, Today's paratransit systems, whose routes have been controlled largely by scheduling software for the past 25 years, rarely make an intelligent adjustment for a "no-show." So much for the pretensions of technology. The point is, a driverless vehicle is hardly a new thing. Given the capabilities of a horse, one could argue that an automobile was, in large part, a regression. The new thing was a vehicle whose brain was inorganic: In other words, a machine. As paratransit service illustrates, the fact that HAV's rely on digital technology is really just a mechanic.
Familiarity and Perspective
One interesting framework in which to view these goings on is art -- particularly Twentieth Century Art. When asked about the similarity of his more-advanced work to a simple child's drawings, Pablo Picasso once remarked that, as his work evolved, he tried continuously to paint things more like a child might see them -- sight as wonderment. A few albums before his death, John Coltrane recorded Om -- a title often wrongly-attributed to some Eastern religious chant. One of Coltrane's former "side men" during his "free jazz" years, with whom I played in the Seventies, explained to me that "Om" (like "Um" or "Uhh") was the universal sound of primitive man -- the sound one might make struggling to push a large rock up a hill. Coltrane was simply searching for the lowest common denominator of human sound: The primal scream. The shrieks and cries that came with efforts to understand the world around us.
In our struggles to figure out why we are evolving toward vehicles we need no longer drive, it might be useful to recall that the first vehicle on which we ever rode was likely a pony -- if not a wooden horse on a Merry-Go-Round. Otherwise, our first cars, buses, trucks, boats, ships and airplanes were just toys. None of them had drivers either.
------------------------------------------------
Note: Special thanks for much of the contents of this article to Robert Pudlewski, Technical Editor of School Transportation News, whose article, "Improving Behavior Through Automation of Vehicle Systems" (STN, February, 2017), summarized both the SAE classification framework and many examples of transitional technologies noted.
The opinions expressed in this article are that of the author and do not necessarily represent the opinions of National Bus Trader, Inc., or its staff and management.
