The National Highway Traffic Safety Bureau estimates that around 40,000 people lost their lives in car accidents last year, with millions more injured. To tackle this issue, automotive and tech companies have been pushing for self-driving vehicles, intending to reduce both the number and severity of crashes.
A significant number of fatal accidents are linked to speeding or alcohol consumption, two things that self-driving cars don’t engage in.
Despite ongoing efforts, the rollout of these autonomous vehicles has hit a snag, primarily due to public concern about safety. A notable incident in 2018 raised alarms, and more recently, the Ministry of Transport introduced a new policy intended to accelerate the deployment of autonomous vehicles. Yet, crucial hurdles remain, like the absence of standardized national safety performance tests.
Just like human drivers can cause accidents, self-driving cars can be involved in incidents as well. However, it’s worth considering that their deployment could still offer significant benefits—but only if they adhere to established safety standards. There’s a clear consensus on improving road safety for everyone.
In December, the Federal Expressway Agency proposed a program named AV Step. Unfortunately, it merely serves as a framework for voluntary reporting without offering concrete guidelines for testing these autonomous vehicles.
The lack of a unified safety framework is likely to hinder public trust and broader acceptance of self-driving cars, particularly when other nations are advancing rapidly in fostering their adoption and safety evaluation.
New drivers, regardless of age, are required to pass basic tests, but autonomous vehicles are not subjected to similar requirements. This discrepancy raises questions about fairness.
Waymo, a company under Alphabet, operates self-driving taxis in cities like San Francisco, Los Angeles, and Phoenix, providing around 250,000 rides weekly. Recently, they’ve partnered with Uber to expand services in Austin, Texas.
In conjunction with Zurich’s insurance giant Swissre, Waymo investigated insurance claims linked to 25.3 million miles driven by their vehicles. The findings revealed an impressive 88% drop in property damage claims and a 92% decline in injury claims.
These statistics underscore the potential public safety benefits of self-driving technology. Yet, according to recent surveys by AAA, only 13% of drivers express confidence in autonomous vehicles. Additionally, a significant majority—about 60%—report feelings of anxiety surrounding them, with prioritization of this technology declining from 18% in 2022 to just 13% this year.
At Mcity, a public-private research partnership at the University of Michigan, efforts are being made to tackle consumer concerns and provide a framework for assessing the safety of autonomous vehicles.
Like the requirements for driver training and competence assessment, both the automotive industry and the National Highway Traffic Safety Administration should back testing protocols for self-driving vehicles on public roads. A two-part methodology has been developed for this purpose.
The proposed “Driver License Test” for self-driving cars evaluates on-road abilities, similar to tests for new human drivers, covering about 50 scenarios. Meanwhile, the Driving Intelligence Test assesses how these vehicles handle complex, high-risk situations.
MCITY is implementing this two-part testing protocol via open-source self-driving vehicle software in its research fleet. By assessing various commercial systems, several operational flaws have also been identified, indicating areas in need of improvement for enhanced driving safety.
The safety assessments don’t have to happen exclusively in Ann Arbor; they can be conducted at suitable locations across the nation to recreate rare and hazardous driving situations.
Successful evaluations will instill confidence in manufacturers, regulators, and the public—ensuring that self-driving vehicles meet essential performance standards for safe operation. Importantly, the testing also protects proprietary AI data for developers.
The promises of autonomous vehicles are huge—they could lead to fewer serious accidents, more efficient delivery systems, and increased access to transportation for those who might struggle with mobility, like seniors or individuals with disabilities.
However, these benefits can only be realized if self-driving cars become commonplace on the roads, which is reliant on consumer confidence in their safety.
