Given the state of vehicle safety in today’s world, driverless vehicles are a promising technology. Since 1946, there have been over 30,000 motor vehicle deaths per year in the United States alone, and during that time there was an 8-year span where deaths exceeded 50,000 people annually.[1] Given these numbers, even small changes to vehicle safety can make a meaningful impact.
Considering a human driver only has two eyes and ears and is prone to distraction, enabling a vehicle with multiple cameras and impressive computing power holds the potential to save thousands of lives each year. However, studies have shown that not only are most Americans afraid of self-driving technology, but some of these fears may be well-founded.
According to a new survey by AAA, 71% of Americans say they are afraid to ride in a self-driving car. AAA asserts that these fears are likely related to the highly publicized deaths in recent years involving driverless technology.[2]
Further, researchers at Georgia Tech have discovered that self-driving AI systems struggle to identify pedestrians of varying skin tones at equal rates. The research suggests that changes during the initial learning may be needed to mitigate what they call “predictive inequity”.[3]
Seeing the impact of vehicle-related personal injury, changes are needed to improve the safety of both drivers and pedestrians. It is also important to keep in mind that new technologies present new challenges that also must be addressed to ensure human safety and right to life.
[1] https://en.wikipedia.org/wiki/Motor_vehicle_fatality_rate_in_U.S._by_year
[2] https://gizmodo.com/71-percent-of-americans-still-dont-trust-autonomous-car-1833284527
[3] https://arxiv.org/pdf/1902.11097.pdf
