Research and development in the field of autonomous driving has seriously stalled after the rapid deployment of electric vehicles.

Manufacturers have preferred to focus their investments around batteries and electric motors, and at the time of writing, technologies related to the autonomous car, or shall we say “semi-autonomous”, are centered around some driver assistance systems that work together.

Lexus is a good student, Tesla and Mercedes are a little less so

Finally, what a job, that’s an important word, especially if you read a recent study by the Insurance Institute for Highway Safety (IIHS). In a report dated March 12, 2024, the IIHS evaluated 14 partial automation systems from various manufacturers. Unfortunately, only a few stood out. Otherwise, it causes concern.
IIHS President David Harkey was blunt in his statements: “Most tested systems do not contain adequate measures to prevent abuse and maintain drivers’ attention on the road. Only Lexus’ Teammate system managed to earn an acceptable rating, while the others were mostly considered mediocre or marginal. » This failure extends to brands such as Tesla and Mercedes-Benz, which, nevertheless, have invested heavily in the development of autonomous technologies.
These deficiencies range from inadequate monitoring of driver alertness to the lack of sufficiently effective reminders to pay attention. Additionally, some systems can be used even if the passengers are not wearing seat belts. It’s a shame for a technology that should provide a plus in terms of security.

Small problems that can be easily solved?

IIHS highlights the insufficient reliability of these systems to guarantee their safety. Despite the potential benefits of semi-autonomous driving, especially on long journeys, there is little concrete evidence that the technology actually makes driving safer. On the contrary, several high-profile accidents have highlighted the risks associated with insufficiently protected systems.
David Harkey urgently calls on manufacturers to strengthen security measures for their systems. He notes that none of the systems evaluated performed satisfactorily in all areas, but he remains optimistic that these shortcomings can be corrected.

According to him, corrections are easily achievable, sometimes just through software updates.