IIHS Reports Pedestrian Detection Tech Rarely Works After Dark

The Insurance Institute for Highway Safety (IIHS) released a study on Tuesday that showcased just how badly advanced driving aids perform at night – specifically the automatic emergency braking systems that are linked to pedestrian detection. 

This mimics earlier studies connected by the American Automobile Association (AAA), which frequently highlighted inconsistencies in driver assistance features in general. But nighttime was when things really started to come undone, with plenty of models failing to register that the simulated pedestrians used for testing were even there. 

Over the last few years, IIHS has been updating its testing its own protocols to focus more on preventative safety technologies. This initially came via updated testing protocols surrounding headlamps but has gradually migrated to other items. The group’s newest testing routine involves simulating how a vehicle’s automatic emergency braking system (the one where the driver does nothing) handles pedestrians crossing the road or walking parallel with traffic.

The results were not good. While the IIHS claimed 19 out of the 23 models tested performed well in daytime conditions, just four vehicles were issued the outlet’s “superior” rating after the sun went down. 

“Eight of the 12 vehicles that earn a basic rating or no credit in the nighttime test got superior or advanced ratings in the daylight evaluation," explained David Aylor, vice president of active safety at IIHS and lead designer of the new testing program.

I’m not going to beat around the bush here. I hate a majority of advanced driving aids because they’ve been grotesquely misrepresented by the industry for years while government regulators, politicians, and insurance groups blindly cry out for more. They’ve also paved the way for increasingly invasive technologies that you couldn’t pay me to install on my vehicle. Your author and just above everyone else that writes for TTAC has at least one anecdote where modern tech went haywire during inclement weather or became so obnoxious while functioning properly that it was deactivated. 

But I didn’t assume that the IIHS would feel similarly. To my great surprise, leadership actually expected a majority of cars to offer subpar performance at night. 

“As we expected, most of these pedestrian [automatic emergency braking] systems don’t work very well in the dark,” said IIHS President David Harkey. “But it’s clear automakers can rise to this new challenge, as Ford, Nissan and Toyota each earn superior ratings for some models.”

Among those tested only the Ford Mustang Mach-E, Nissan Pathfinder, Toyota Camry, and Toyota Highlander earned top marks. Though the Honda Accord, Hyundai Palisade, Hyundai Sonata, Nissan Frontier, Nissan Murano, Subaru Ascent, and Subaru Outback still managed to earn “advanced” ratings. The rest all earned “basic” scores or tested so poorly across the board IIHS deemed them unworthy of any credit. 

”It’s discouraging that so many midsize SUVs and small pickups perform poorly in the nighttime test because research suggests these types of vehicles are more dangerous to pedestrians,” he added. 

One of the reasons the outlet is so keen on pedestrian safety is because deaths have been on the rise for a while and primarily take place at night. The first person struck and killed by an autonomously driving vehicle also took place at night, perhaps underpinning just how important it is to make sure these systems function after dark. 

Then again, maybe the real issue is that drivers are dropping the ball and modern safety technologies don’t make a great Band-Aid. We’ve beaten this expired horse before, though it bears repeating. There is a massive amount of evidence that advanced driving aids not only have trouble working as promised but also encourage motorists to take a more passive role behind the wheel. Remember in the above paragraph where I mentioned the first person killed by an allegedly autonomous vehicle driving at night? Well, the safety driver of that Uber vehicle was watching a video on their phone, totally disengaged from the road, when it happened.

However, that’s not how the IIHS sees things, as it’s championing a recent study (it conducted) claiming that automatic braking reduced pedestrian crashes “by more than a quarter overall for equipped vehicles.” Unfortunately, no improvements were noted after sundown. So, from the perspective of insurance groups, the issue isn’t that the technology kind of sucks and dulls the senses. It’s that pedestrian detection simply doesn’t work well enough when it gets dark outside. Be that as it may, IIHS still says it wants to see these and similar systems installed into more vehicles in the coming years. 

Though it seems to be setting a fairly low bar for excellence. Only Nissan’s Pathfinder avoided a collision with the pedestrian dummy in both test scenarios at all test speeds with both its low and high beams. Every other car (including those that got high marks) struck a simulated person in at least one of the scenarios. These included a faux pedestrian crossing test conducted at 12 mph and 25 mph, followed by the faux pedestrian walking parallel with traffic conducted at 25 and 37 mph. These events were also repeated to contrast whether or not high or low-beam headlights made any difference.

Testing protocols also made some exceptions for nighttime driving. During the daytime sequence, there was a third scenario where a child-sized dummy pops out between cars. But the IIHS nixed it “because few child pedestrian fatalities occur at night.” However, AAA’s earlier studies showed this to be one of the hardest tests for any vehicle to pass, presumably meaning lower scores across the board in the IIHS study if it had been included.

[Image: Ryan DeBeradinis]

Become a TTAC insider. Get the latest news, features, TTAC takes, and everything else that gets to the truth about cars first by  subscribing to our newsletter.