Thermal imaging offers an affordable night vision solution that satisfies NHTSA requirements for nighttime AEB systems. Beyond night time emergency braking, integrating thermal imaging into the sensor suite enhances vehicle perception for both day, night, and low visibility conditions (rain, fog, or glare). Fusion of visible and thermal images improves robustness against false positives and spoofing scenarios where a visible camera alone may misclassify objects, preventing phantom emergency brakes. As a passive technology, thermal cameras detect heat emission without relying on illumination, making them immune to disturbances from headlights, RADAR, or LIDAR from surrounding vehicles. We propose a windshield architecture that co-integrates visible and thermal cameras, and we support its benefits with quantitative analysis from field trials and a publicly released multimodal dataset covering diverse weather and lighting conditions. This dataset aims to accelerate the development of perception models that leverage thermal vision for reliable ADAS.