Driverless cars – where standards intersect

The promise of driverless cars continues to ramp up, with tests on public roads occurring or planned all over the world.

I’ve spoken before about the emergence of this technology, specifically on the importance of standards in governing how autonomous vehicles communicate with one another. Whether they retain their own vehicle-to-vehicle (V2V) standards or are absorbed into the larger network promised by the Internet of Things (IoT), standards will play an enormous role in the technology’s success or failure.

However, there is another set of standards that may need to be addressed, and it’s one that could have a tremendous impact on not only how driverless cars behave, but the impact they have on traffic and congestion.

Traffic Jams – A Constant Standard?

Less traffic, more comfort and increased safety are often touted as prime benefits of autonomous vehicles, but how well do those two goals truly complement one another? For instance, trains are conducive to comfort and safety in part due to their slow rate of acceleration and deceleration. Travelers often study or do work en route, not least because the ride is less jerky than a car. To maximize comfort and safety, driverless cars would need to mimic that behaviour.

A recent study at Imperial College London examined this very issue. They simulated a basic four-way intersection where 25% of the vehicles were driverless, with the remainder human-operated. Sixteen scenarios that varied acceleration speed, following distance and length of yellow lights ran 100 times each for an hour.

In every case, traffic worsened against a baseline of all human drivers. Fewer cars passed through the intersection per cycle, causing significant backups and congestion.

Driverless cars may very well hold the promise of a comfortable drive, efficient traffic and less congestion – just not all three at the same time.

Determining a Standard Drive

If the journey to developing a data standard that enables vehicles to communicate has been slow (a European standard in the works since 2009 only saw release in 2014), then determining a standard for how they drive could be even more difficult. Who will decide what kind of driving is appropriate driving? Will one region decide to optimize traffic speed and efficiency while its neighbour opts for safety and comfort? Will passengers have any control whatsoever? Moreover, how will vehicles know that they have exited one zone and entered another if the data standard isn’t designed to accommodate local variances?

The interaction between driving behaviour and communication standards could become one of the most important issues to face driving since the introduction of mandatory licensing in the 1900s.

We’d love to know your predictions on driverless cars, particularly where driving behaviour is concerned – visit the CSIO eXchange discussion forum to join the conversation.