Analysis by Daniel Stern, DVN Chief Editor
Four years ago, back in 2021, the U.S. Congress passed an infrastructure bill into law which, among many other things, ordered NHTSA to permit ADB as specified in SAE J3069; to ensure that headlamps are correctly aimed on the road, and to ensure that systems are tested on-vehicle. And DVN analyzed those congressional orders (article1, article 2).
Now, NHTSA have published a new study flowing from that 2021 set of congressional directives. It’s called ‘Lower Beam Headlighting System Performance On-Vehicle Test Procedure Examination’ (DOT HS 813 702), and it describes NHTSA’s work to develop a candidate test protocol for evaluating the performance of low-beam headlights installed on the vehicle, rather than isolated (just the headlamps, on a photogoniometer).
It’s based on an earlier NHTSA draft protocol and work by UMTRI (the University of Michigan Transportation Research Institute). Low-beam performance was determined by switching on the lamps as installed on a production vehicle and measuring the amount of light cast forward onto the roadway. Illuminance was measured over an array of locations determined by translating FMVSS 108 low beam test points from polar coordinates to lateral, longitudinal, and vertical test locations on a flat roadway.
Seven vehicles’ headlights were tested this way, to collect data on the protocol’s accuracy and repeatability. The on-vehicle measured illuminance averaged 7 per cent higher than illuminance values measured using regular lab-based methods. The difference between the illuminance measured in lab testing versus on-vehicle testing varied greatly by test vehicle and measurement location. There was large relative difference between on-vehicle results and the lab-test results for many test vehicles and test point combinations. The relative difference ranged from -311 per cent to +80 per cent. There was great variance in relative difference from test vehicle to test vehicle, and from test point to test point. The average relative difference, over all test vehicles and measurement locations, was -35 per cent. The absolute average relative difference, calculated over all test vehicles and measurement locations, was 48 per cent.
To assess the repeatability of the protocol, the full test procedure was performed five times on a single vehicle. Results showed that measured illuminance values near the cutoff had the greatest variability. Overall, this work revealed the challenges of using an on-vehicle measurement method rather than a lab-based method; the on-vehicle measurement method was not as accurate as the lab-based method.
These results point out (and practically shout!) out the reason why lab-based measurement is the longstanding, globally-agreed way to collect the data to accurately and repeatably characterize the performance of headlamps. Lab testing cuts out the numerous variables and influences that affect the light distribution once the lamps are on a car. Michael Hamm scrutinized those variables and influences in 2020. These new NHTSA findings of imprecision and inaccuracy in trying to measure headlamp aim on-vehicle rather than in-lab confirm Hamm’s findings: the on-car aim is highly variable, even if the lamps are carefully adjusted to the nominal aim setting.
And that’s at the root of a lot of the problems caused by poor lamp aim (too much glare and not enough seeing). The congressional directive to pursue vehicle-based rather than lab-based lamp aim measurement was probably put forth with good intent, but it was a misguided directive, probably based on inadequate technical expertise. That is not terribly surprising, given the tacked-on/snuck-in appearance of the “headlights” section of the infrastructure law, in amongst all the rest of the law’s coverage (roads, bridges, etc). It’s consistent with how these kinds of things tend to work in the U.S. lawmaking system. But it would have been better if Congress, instead of baselessly declaring that vehicle-based aim measurement is the thing to do, had ordered NHTSA to find ways of reducing the divergence between intended and actual lamp aim. That would have paved the way for automatic headlamp aiming systems (or at least auto-levelling systems, as a half-step in the right direction).
NHTSA also released another study, ‘Vehicle Automatic Lower Beam Activation System Test Procedure Development’ (DOT HS 813 683). This describes NHTSA’s development of a draft test protocol to measure the performance of automatic ambient-light-sensitive headlamp activation systems. The goal of this research was to understand the current state of automatic headlamp activation system performance and develop a way to measure system performance. The objectives were to develop a draft test procedure for measuring the ambient light level and response times of automatic low-beam activation and deactivation (how long after the stimulus light level is reached does the system respond?) and to measure these values with the draft test protocol for a small set of test vehicles.
The developed draft test procedure drew from UN R48 (the version current in 2016). A lamp with adjustable light level setting provided artificial ambient light on the tested systems’ light sensors; a laboratory-based method was devised for simulating ambient sunlight coming from a realistic direction and measuring that ambient illumination at a vehicle’s exterior windshield surface.
Five 2022-model passenger cars were tested with the draft protocol; results showed a range of values for both the ambient illumination level values and response times for headlamp activation and deactivation. Most measured values met UN R48 requirements. The draft test procedure was also performed on one of the test vehicles a second time in outdoor, natural sunlight conditions; comparison with lab results showed good correlation. Overall, the draft test procedure was found to be useful for assessing performance of vehicle automatic lower beam headlamp activation systems.
It is interesting to see this work from NHTSA now. It reminds that eight years ago in 2017, we reported on Transport Canada’s adjustments to CMVSS 108 to help eliminate “ghost cars” (driving after dark with only daytime running lights). One of the three countermeasures was R48-style automatic low beam switching.
