Original text here from Patrice Bernard (LinkedIn)
After a period of initial excitement, doubts have recently surfaced in the world of autonomous driving. A key player in this field is conducting numerous scientific studies to show that autonomous driving is still a major factor in the improvement of future automobiles, particularly in reducing accidents.
This skepticism is not surprising for a technology that aims to replace human intelligence in one of our most common and daily activities. However, the debate over the quality of computer systems gradually taking control of driving is skewed. A few incidents, quickly broadcasted worldwide, are enough to question their reliability, overshadowing the significant damage caused by human drivers.
To rationalize these concerns, Waymo, which operates driverless fleets in two major American cities, has partnered with reinsurer Swiss Re. Last fall, Swiss Re published a groundbreaking analysis objectively comparing its own liability claims data from all currently operating vehicles with incidents during Waymo's autonomous trips.
The initial findings, set to be published in a peer-reviewed journal, are striking. They confirm the safety benefits of autonomous vehicles. Once adjusted for factors like mileage and geographic distribution, autonomous driving software showed a 76% reduction in the frequency of material damage and a 100% reduction in personal injuries (with no cases recorded).
Subscribe to our newsletter:
The differences seem easily explainable. Despite the potential for coding errors and other unforeseen circumstances, and excluding absolutely unavoidable accidents, the robotic driver has an undeniable and unmatched advantage over human drivers, who are often the cause of most accidents due to inexperience, distractions from phones or passengers, fatigue, alcohol abuse, or other substances.
The enthusiasm for these results might be tempered by the statistical significance of the study. Although its foundation is solid, it's based on just over 6 million kilometers driven by Waymo vehicles, which would equate to only about four instances of injuries according to estimates based on 600,000 claims and 200 billion kilometers covered in the insurance data. Thus, the performance needs to be confirmed over time.
Swiss Re's initiative, in collaboration with Waymo, is unusual but logical. Even though the development of autonomous vehicles is still in its early stages, it's likely inevitable and will have a significant impact on insurance models developed over decades of incremental progress. Beyond new liability challenges when human drivers are phased out, it's crucial to accumulate reference points to adjust actuarial calculations for a radically different generation of vehicles.
