What we will see in the shorter term is increasingly sophisticated vehicle automation
by Peter Vogel
There is a revolution underway that is already having an impact on how we plan for the future of transportation. It is a revolution with broad implications across many sectors of the economy.
At some point, perhaps within as few as two decades, one consequence of this revolution will be that those in their late teens will no longer need that rite of passage, a driver’s licence.
Yes, we are talking about self-driving vehicles, or autonomous vehicles. Vehicular automation or semi-automation if you will.
We are many years away from completely autonomous vehicles – vehicles that require no human interaction, save for the provision of a destination. Somewhat sooner we will see autonomous vehicles on tightly controlled highway segments, such as freeways.
What we will see in the shorter term is increasingly sophisticated vehicle automation through the use of mechatronics, artificial intelligence, and various systems to aid the vehicle operator.
At its simplest, we are already seeing this in many current car models that feature the likes of lane-change warnings when there is an object in the operator’s blind spot, forward brake warnings when a system detects a rate of change of closing distance beyond a certain threshold, and adaptive cruise control that adjusts according to traffic patterns.
However, these sorts of technologies leave us a long way from truly autonomous cars, vehicles that are completely aware of their environment and that can make decisions to get said vehicle from point A to point B without any driver input.
Such vehicles need much more sophisticated technologies, such as computer vision, various forms of distance-sensing radiation, for instance radar and lidar (light-radar), all with the intent of interpreting the vehicle’s surroundings, be it other vehicles, people, and road signage.
For instance, a truly autonomous vehicle would need to be able to distinguish between a police officer attempting to wave down a vehicle and a person waving randomly at a passerby.
In 2014, SAE International, a body that develops standards for transport industries, including automotive (perhaps you recognize the acronym from a can or bottle of car oil), established a series of levels to categorize vehicular automation.
On the SAE scale, the lowest entry, Level 0, essentially describes a vehicle that can provide warnings, but for which there is no intervention or attempt at vehicle control.
At the highest level, Level 5, no human interaction beyond provision of a destination is required. The vehicle is capable of reaching any destination accessible via legal roads completely on its own.
Clearly, Level 5 automation is not something we will see in general use for many years. Such automation requires inter-vehicle communication; in other words, cars that talk to each other and that make decisions for the common good.
Most new vehicle automation will be focused on Levels 2 and 3. At Level 2, the vehicle systems control acceleration, braking and steering, but the driver must be prepared to take over at any time. At Level 3, the driver can work on other tasks when the vehicle is operating in restricted environments, such as a freeway, but must be prepared to take over if needed.
Level 2 automation is the designation given to the Autopilot mode in the Tesla S. It became the object of litigation following the death of a driver when his vehicular systems could not distinguish a white semi-trailer and truck from the bright sky. The Tesla S was in Autopilot mode at the time of the crash.
In January, the US National Highway Traffic Safety Administration (NHSTA) found that the operator had ignored Tesla’s guidance that control be maintained at all times, even when in the automated mode.
A contrary finding by the NHSTA in this case would have sent quite a chill through the nascent vehicle automation sector. Since the accident, Tesla has incorporated an auto-cut-off into the Autopilot mode if the sensors detect that the operator isn’t paying attention.
In the long run, vehicular automation will have enormous benefits, such as dramatic reductions in collisions, and, by extension, huge savings in medical costs. There will be better traffic flow and reduced fuel consumption. Presumably, insurance rates will eventually decline markedly as well.
And those are just some of the impacts we can expect from vehicular automation.
Follow me on Facebook (facebook.com/PeterVogelCA) and on Twitter (twitter.com/PeterVogel).