You’ve certainly heard how autonomous cars have been touted in recent years. They have been presented as an effective way to limit d’1 ds to traffic accidents. However, a recent study was published last Thursday, telling us that these autonomous cars could probably only prevent a third of all road accidents in the United States. The researchers found that the remaining accidents cannot be prevented by these cars since they are caused by errors unable to be caused by autonomous driving systems. This study was conducted by the Insurance Institute for Highway Safety (IIHS), a research group funded by U.S. insurers. In order to overestimate the number of accidents that could continue to occur if autonomous cars are made to make the same decisions as humans, the researchers analyzed more than 5000 accidents. These accidents are those that were reported by the police as part of the national survey on the causes of motor vehicle accidents.
According to the study’s calculations, 72% of accidents analysed were vitable if those caused by speed lexcs and traffic violations were taken into account. However, some remained inviting, such as those caused by a vehicle failure such as a ruption or axle. The researchers were able to spare the factors read to the driver that contributed to the accidents. These are mainly errors in decision-making, forecasting and performance, as well as errors due to driver’s incapacity.
Keep in mind that a self-driving system will not correct these errors automatically. To illustrate this, the researchers have quoted the example of the fatal collision of an Uber vehicle with the pitonne Elaine Herzberg in 2018. It is remembered that the car’s autonomous driving system had not only had difficulty initially detecting Herzberg, but had also not been able to predict where she was going and therefore had not performed the quick manoeuvre that would have saved her life. The conclusion made by the IHS is that the accident rate would decrease only if autonomous cars gave priority to safety over drivers’ concerns. This is what says Alexandra Mueller, IIHS research scientist and lead author of the study: Our analysis shows that it will be crucial for designers to prioritize driver safety if autonomous vehicles are to deliver on their promise to be safer than human drivers.
Once all these essential points that are the causes of accidents have been well determined, it is up to the designers to ensure that these autonomous cars provide solutions. Otherwise, they won’t really change things. This is clearly what Jessica Cicchino, IIHS Vice-President of Research and co-author of the study, said: Our goal was to show that if you do not address these problems, autonomous cars will not bring safety benefits standards.
How about you?
How did you feel about the conclusions of this study?
A Volkswagen executive says Level 5 autonomous cars (fully autonomous driving) may never arrive. However, Level 4 seems to be ralegible to him
People who know more about autonomous driving are more confident about technology, according to an survey
An Apple ingnieur who died in the crash of his Tesla vehicle was d’dj complained about autopilot software and its malfunctions, after the application of an update