IIHS Put a Safety Vest on a Crash Dummy. Two Out of Three Cars Didn't Slow Down.

Honda's CR-V registered zero percent speed reduction from its pedestrian automatic emergency braking system when the crash dummy wore retroreflective strips on its limbs and joints. Not in the dark. Not in dim lighting. In every lighting condition IIHS tested.^[1]

Those same strips that make a construction flagger glow like a lighthouse to your eyes make them a ghost to the camera behind your windshield. IIHS tested three 2023 SUVs against a pedestrian dummy dressed in four different outfits, including the retroreflective patterns that OSHA requires for roadway workers. CR-V and CX-5 hit the dummy in 84% and 88% of all test runs, respectively. Subaru's Forester stopped in every trial but one.^[1]

That is not a typo. America's best-selling compact SUV^[2] cannot distinguish a road worker in a safety vest from the road itself.

Context makes it worse. In 2024, 850 people died in work zone crashes; 156 were pedestrians, mostly construction and utility workers standing feet from live traffic.^[3] Bureau of Labor Statistics data shows 54 worker-pedestrians struck and killed in work zones annually.^[4] Every one was probably wearing employer-mandated reflective gear. And every one was likely invisible to whatever Honda or Mazda AEB system was bearing down on them.

Subaru's dual-camera stereo system handled every outfit, every light level, stopping cold in nearly every trial. Meanwhile, the CR-V's camera-and-radar setup didn't flinch at the reflective jacket either: zero percent speed reduction, zero conditions, zero exceptions. Just a full-speed impact.

IIHS senior research scientist David Kidd put it plainly: the moving reflective strips that help human drivers recognize a walking person "probably confounded" the AEB sensors.^[1] A feature designed to enhance human perception actively sabotaged machine perception.

IIHS only tested three vehicles. Nobody knows how many others share the blind spot. Honda's CR-V fleet alone is 3.15 million vehicles.^[2] Honda and Mazda combined account for roughly one in ten cars sold in the U.S. If their AEB can't see reflective vests, a substantial fraction of the nation's crash-prevention fleet works better when the pedestrian is dressed in black.

IIHS President David Harkey called it "untenable." That's diplomatic. A more accurate word is backwards. We built a safety protocol where workers wear reflective gear to be seen by drivers, then automated the driving and forgot to test whether the automation could see the gear. Now the protocol and the technology are working against each other, and the people caught in the middle are the ones standing next to the orange cones at 2 a.m.

Fair counterargument: pedestrian AEB still cuts crash rates by 27% overall, and Subaru proved the problem is solvable.^[1] Honda and Mazda could push a software update tomorrow. But until they do, the gap is live. FARS doesn't track whether a victim wore a safety vest, and NHTSA doesn't require AEB testing against retroreflective clothing. Nobody is counting the workers this blind spot has already cost.

If you drive a Honda CR-V or Mazda CX-5: your pedestrian AEB may not detect workers in safety vests, especially at night. Do not rely on it in work zones. Slow down manually. If you're a road worker: the vest helps human drivers see you. The car behind the human may not agree.

Sources & References

1. IIHS, “High-visibility clothing may thwart pedestrian crash prevention sensors,” January 9, 2025. iihs.org
2. NHTSA, Fatality Analysis Reporting System (FARS), 2014–2023. nhtsa.gov
3. NSC Injury Facts, “Work Zones,” 2024 data. injuryfacts.nsc.org
4. Bureau of Labor Statistics, Census of Fatal Occupational Injuries (CFOI), worker-pedestrian fatalities in work zones, annual average. injuryfacts.nsc.org