Back Industry News

How Much Safer Should Self-Driving Cars Be? Try 0% Posted on Jul 10 - 2018

Share This :

Two car accident victims stumble into a bar. The first tells the bartender, "it came out of nowhere." The second has already sent pre-crash video footage and telemetry data to the bartender before arriving.

Rhetorical arguments are bubbling up around what it means for a self-driving vehicle to be safe enough for commercialization, and yet, statistical projections overlook the core value propositions of driverless technology: observation and communication. In short, autonomous cars transform accidents into errors, which is why they stand to be the greatest safety innovation in automotive history -- even if they don't prevent a single crash.

The U.S. saw 7 million traffic collisions in 2016. This is a key performance indicator for benchmarking safety, but our inability to observe human drivers means the figure only reflects crashes reported to authorities. An extensive survey from the NHTSA itself found that drivers admitted failure to report their collisions at a rate of 29%. As such, flipping the switch on autonomous cars nationwide and witnessing a 0% improvement in reported crashes would actually indicate a 29% decrease in real crashes -- the most significant improvement in history, thanks entirely to observation.

Such a monumental advance takes a step further when considering any collisions that would occur:

Consider that the paramount safety feature in a conventional vehicle is its driver, but the driving public has no reliable model for learning from individuals' mistakes. Nowhere is this flaw more apparent than in the comparatively astronomical fatality rates for young drivers, who are three times more likely to die than drivers over twenty years of age. Even within the subset of teenage drivers, 16-17 year-olds are twice as susceptible as 18-19 year-olds.

In contrast, autonomous vehicles at large would immediately create a more equitable age distribution of crash victims -- a small moral victory in its own right. But more importantly, fleets of self-driving cars gain driving wisdom in unison, and therefore break humanity’s continuous cycle of reintroducing dangerous novice drivers to the traffic mix. Through communication, driverless tech becomes a single cohort that only improves with age, as opposed to suffering a never-ending influx of ignorance.

The wisdom bestowed by observation and communication would also extend beyond vehicles themselves. By replacing all human drivers with a handful of software programs, the most complicated variable of a crash assessment would be marginalized. A year spent enduring a 0% improvement in collisions could yield more valuable data on road design flaws, mechanical faults and maintenance needs than has ever been collected previously. View More

x

Get the Global Big Data Conference
Newsletter.

Weekly insight from industry insiders.
Plus exclusive content and offers.