'A production autonomous vehicle that can go anywhere, anytime isn't available at your local car dealer and won't be for quite some time. We aren't there yet.'

The current crop of semi-autonomous driver assists aren't proper substitutes for hands-on attention, according to a new study from the US Insurance Institute for Highway Safety (IIHS).

According to the non-profit's research, current semi-autonomous systems like lane-keeping assist and adaptive cruise control displayed behaviours ranging from "the irksome, such as too-cautious braking, to the dangerous".

Five cars were put through their paces as part of the tests: a BMW 5 Series, Mercedes-Benz E-Class, Volvo S90, Tesla Model 3 and Tesla Model S. Their adaptive cruise control systems were put through their paces on a track, with the goal of finding out how they handled stopped cars, cars leaving their lane, and the way the accelerate/decelerate.

The first test involved driving at 50km/h towards a stationary car with adaptive cruise turned off, to test the effectiveness of each autonomous emergency braking system. Three of the five cars stopped – only the two Teslas crashed into the stopped dummy car.

With adaptive cruise enabled, all five stopped, although the researchers noted Tesla's system was earliest to brake and smoothest to stop. The Volvo, meanwhile, pulled 1.1G in deceleration, having left its braking to the last minute.

Finally, the testers had a 'lead car' pull out of the lane, revealing a stationary vehicle on the road. All five vehicles stopped, even though the brands suggest maybe they won't. During subjective on-road testing, researchers found instances where each of the cars failed to detect a stationary vehicle, requiring the driver to intervene.

When it came to lane-keeping, the testers recorded how the cars reacted to hills and curves, recording how often the cars touched a line, crossed a line, disengaged their system and stayed correctly within their lane. The results are below, but here's the general gist: the Tesla Model S and Model 3 performed very, very strongly.

"Designers are struggling with trade-offs inherent in automated assistance," said David Zuby, IIHS chief research officer.

"If they limit functionality to keep drivers engaged, they risk a backlash that the systems are too rudimentary. If the systems seem too capable, then drivers may not give them the attention required to use them safely."

"We're not ready to say yet which company has the safest implementation of Level 2 driver assistance, but it's important to note that none of these vehicles is capable of driving safely on its own," Zuby went on.

"A production autonomous vehicle that can go anywhere, anytime isn't available at your local car dealer and won't be for quite some time. We aren't there yet."