comment on this article

Self-driving cars and the challenges they face to gain acceptance (5/6)

The nature of wheeled transportation is changing very rapidly in what has become a disruptive technological environment.

Radical upheavals are taking place along several different paths. The three key ones are in relation to the powertrain (with the migration away from internal combustion engines and toward electric propulsion), the widespread proliferation of x-by-wire (thereby removing heavy and less reliable mechanical systems) and finally – most importantly of all – who/what will actually be responsible for driving future generations of automobiles.

Regardless of which form of motive force is used under the hood, the constituent electronics hardware and the roles that this is taking in vehicles’ management and operation continue to grow at a considerable pace. With this, the degree of autonomy that automotive systems possess is increasing and, consequently, driver involvement is set to diminish in the years ahead.

While the previous two blogs in this six-part series focused on the supporting technology – sensors and communication infrastructure respectively – the next two will be concerned with how to bring about its wider adoption, and where there could be potential flashpoints.

One aspect of intelligent vehicles that promotes safer traffic conditions is their interconnectivity. Cloud-based systems can not only connect one car, bus or truck with another, but can also connect all of them to a municipal traffic management system. This interconnectivity will help ensure cars are aware of traffic lights, potential obstructions to their route and other road users – thereby reducing the burden placed on simple sensor recognition, and giving a more comprehensive overview of what is happening at any point in time.

It should be noted that it is not just a matter of the position, speed and direction of travel pertaining to vehicles in the vicinity. Road etiquette will have a significant influence on the kinds of decisions that self-driving cars make. City driving often requires a degree of assertiveness than would not be normal when on rural roads. It frequently involves manoeuvres – such as merging into a busy lane – that to an autonomous vehicle may appear too risky to undertake.

In cultures where drivers in the receiving lane typically slow down to let in another vehicle out of politeness, it will be much easier for the software to decide the level of risk is acceptable and go ahead with the merge. In other parts of the world, however, drivers may not give way so easily. Under such circumstances, the artificial intelligence (AI) software will need to take on a much greater level of risk – pulling into the lane in the expectation that the driver approaching from behind will realize and brake quickly enough.

As the early days of autonomous driving will be in environments where machine-based and human drivers coexist, collisions seem inevitable for a variety of reasons. It will not be until much farther into the future that vehicles will be able to negotiate with each other wirelessly and determine who should join the lane based on a queueing algorithm.

Another question now being asked is whether road etiquette will be affected by self-driving cars. Studies have already shown that the presence of autonomous vehicles being tested on our roads can change the way drivers in the vicinity behave – more specifically, drivers tend to show a more aggressive and less considerate attitude. Is this purely down to the novelty value of there just being a handful of trial vehicles in operation currently, or is there something more deep-set within the human psyche that needs to be investigated?

There have even been cases where individuals have actually been witnessed using handguns to take pot-shots at self-driving vehicles while they were in the process of traveling around their test routes. The motivation for such incidents is unclear. It could be anger about the possibility of autonomous technology leading to people losing their jobs (taxi drivers, haulage drivers, etc.), or maybe just an instinctive aversion to the unknown.

It is perhaps naïve to think that society’s willingness to embrace autonomous driving is unlikely to be an amorphous thing. Some demographics are going to be more accepting of it than others, and age (in particular) is certain to be factor. Those who have been driving for several decades are probably not going to be as easily persuaded to consider trying a mode of transport that seems completely alien to them (and where they don’t have full control) as someone who has only just passed their driving test. Initiatives are now underway in several European countries to address this – making sure that older generations’ concerns are not ignored, and attempting to encourage their engagement.

By the very nature of AI, the algorithms on which AI inferences are based will need to be programmed by humans, and this, in turn, raises a number of dilemmas. There is already heated debate about the whole decision-making process that autonomous vehicles will need to follow when put in life-critical situations.

Questions of this kind have been asked ever since people started handing responsibility for social order to third parties. In this context, law, municipal traffic management, pedestrian safety and myriad other issues pertaining to how vehicles should interact with society will eventually be placed in the hands of solid-state decision-makers (though these will draw on a framework that human engineers have bestowed on them).

This responsibility doesn’t rest solely with the vehicle, of course. Autonomous transportation is just one part of the all-encompassing smart city ecosystem. An urban area’s cloud infrastructure will be as critical to traffic safety as the state of the road network or weather conditions. Sophisticated vehicle-to-everything (V2X) network systems must be set up. They will need to be scalable, as well as ultra-reliable and responsive. That burden falls on the engineering community to create the components and subsystems required. At the same time, there are a plethora of fundamental questions that need answering when it comes to who will be accountable if something does go wrong.

These will be addressed in the final part of this blog series.

Mark Patrick, Mouser Electronics

Comment on this article

This material is protected by MA Business copyright See Terms and Conditions. One-off usage is permitted but bulk copying is not. For multiple copies contact the sales team.

What you think about this article:

Add your comments


Your comments/feedback may be edited prior to publishing. Not all entries will be published.
Please view our Terms and Conditions before leaving a comment.

Related Articles

A new era

How can 5G help to unlock the potential of the Industrial IoT?


Blockchain users are rediscovering hard lessons from real-time systems and game ...

Wireless monitoring

From wood and charcoal for providing heat, to dietary animal fats and grain for ...