Teledyne FLIR is Saving Lives with Pedestrian AEB

Teledyne FLIR recognizes the need for thermal nighttime testing protocols. In September 2024, Teledyne FLIR collaborated with VSI Labs, the leading operator of ADAS/AD Development Vehicles used for the test, evaluation, and assessment of ADAS and automated driving components.

VSI Labs integrated the latest Teledyne FLIR LWIR camera module along with associated IR pedestrian detector into their research vehicle. The testing was performed in accordance with the Final Rule of the Federal Motor Vehicle Safety Standard (FMVSS) No. 127 that require automatic emergency braking (AEB), including pedestrian AEB (PAEB), systems on light vehicles.

At AutoSens Europe, Teledyne FLIR presented the analytics and results from this testing, integral to increasing pedestrian safety and saving lives.

John Eggert, Global Head of Business Development for Automotive at Teledyne FLIR, discusses how thermal imaging technology is advancing pedestrian automatic emergency braking (AEB), with a focus on improving nighttime safety and meeting new regulatory standards.



John Eggert, Global Head of Business Development, Automotive, Teledyne FLIR

Can you explain the importance of infrared (IR) technology in enhancing pedestrian detection, especially during night-time driving?
The new U.S. regulations mandate that starting in 2029, all vehicles sold must be able to detect and stop for pedestrians in complete darkness. Vehicles must stop for a pedestrian at speeds of up to 72 kilometers per hour. This is a major shift from current regulations worldwide, even compared to Europe's advanced standards under Euro NCAP, which simulates street lighting conditions. If automakers can't meet this standard by 2029, they won't be able to sell their vehicles in the U.S.

Until now, pedestrian detection systems have relied mainly on visible cameras and radar. However, these sensors have limitations in certain conditions: visible cameras struggle in low light, and radar, while reliable in all weather and lighting conditions, provides only low-resolution data. Radar also struggles to detect stationary or slow-moving humans. Therefore, alternative technologies like thermal imaging are emerging as potential solutions because they work well in the dark.

The new FMVSS No. 127 rule mandates PAEB systems by 2029. How do you foresee the role of thermal imaging technology evolving in meeting this regulatory deadline?
We have conducted tests that show thermal imaging can detect pedestrians at long distances, up to 200 meters, meeting the new Federal Motor Vehicle Safety Standards (FMVSS) with a significant safety margin. The choice now lies with automakers on what combination of sensors—thermal imaging, LiDAR, visible cameras, and radar—provides the best performance and value. Thermal imaging stands out because it offers better performance than existing camera and radar setups and is more affordable than LiDAR.

How does Teledyne FLIR’s LWIR camera module integrate with other sensors like radar and visible cameras to improve the accuracy of automatic emergency braking (AEB) systems?
We have tested a setup that integrates thermal cameras with visible cameras and radar. This combination provided good results, outperforming standard vehicles that use only visible cameras or radar. In our tests, vehicles equipped with thermal, radar, and visible cameras passed all requirements easily. In some cases, the thermal and radar data alone were sufficient for accurate detection. The thermal camera identifies the object as a person, while the radar provides precise distance measurements, allowing for better planning and control of the vehicle's braking.

Could you share some insights from your collaboration with VSI Labs on thermal night-time testing protocols, and how the results are expected to influence future pedestrian safety regulations?
VSI Labs assisted with vehicle integration and data analysis for our tests. The results showed that thermal imaging performs well beyond the FMVSS requirements. For instance, we tested scenarios with child-sized pedestrian dummies and pedestrians lying on the ground, which are more challenging than the standard tests. We believe these tests could influence automakers and regulatory bodies to adopt more comprehensive safety measures.

What are some of the key challenges in developing thermally active pedestrian test mannequins, and how is Teledyne FLIR working to address them?
Traditional pedestrian dummies used for testing aren't designed to emit heat, which is critical for thermal imaging. Since humans emit body heat, we need dummies that simulate this thermal signature. The industry is now adapting by developing heated dummies that better represent human thermal characteristics. We used such heated dummies in our recent tests, and we expect demand for these advanced test dummies to grow as more automakers adopt thermal imaging technology.

Looking ahead, how do you see thermal camera technology advancing in the context of automated driving systems beyond pedestrian detection?
Thermal cameras can detect more than just pedestrians; they are also effective at identifying other vehicles on the road. For example, self-driving trucks need to detect vehicles hundreds of meters ahead to plan lane changes or adjust speed, and thermal imaging excels in these long-range detection scenarios, even in low visibility conditions like darkness or rain. We believe thermal cameras will play a significant role in enhancing safety for autonomous vehicles.

www.auto-sens.com