How Advanced Crash-Avoidance Tech Will Change Car Safety by 2029

As we approach the September 2029 compliance deadline for the National Highway Traffic Safety Administration’s new FMVSS No. 127 rule mandating new vehicles be capable of detecting other vehicles, objects, and pedestrians and then providing automatic emergency braking, the agency’s updated New Car Assessment Program will reward advanced crash-avoidance technologies in hopes of encouraging early adoption and improvement in sensing—especially after dark.

Not surprisingly, then, several novel sensing upgrades were unveiled at the 2026 Consumer Electronics Show. Teradar works like radar, but by emitting higher-frequency (terahertz) electromagnetic waves, angular resolution (and hence point-cloud clarity) is said to improve to nearly that of lidar at lower hardware cost. Meanwhile, Microvision is pursuing distance- and velocity-sensing frequency-modulated continuous-wave lidar via Scantinel’s silicon-photonics printed circuit board approach, dramatically lowering its cost. Then Atomathic showed off AIDAR, which marries inexpensive radar hardware with artificial intelligence to better interpret the vast amount of seeming gibberish arriving back at the unit from signals that may have bounced off multiple surfaces—allowing it to discern, for example, a low-reflectance human standing next to a highly reflective delivery van. This all leverages radar’s lower hardware cost.

Interior sensing is poised to make big leaps forward, as well, using better cameras and image processing for driver monitoring and/or occupant detection and rear-seat reminder tasks. Aptiv just introduced such a system that can determine occupants’ size, shape, posture, and seat belt usage, then classify each person by size/type and tailor the restraint deployment accordingly. It claims to be the first camera system to do so with sufficient fidelity to eliminate all other costly occupant sensors in seats and on seat tracks. TrinamiX reckons that by adding an infrared dot-matrix projector to the camera and sensing its reflections, it can measure heart and respiration rates, secure face authentication, identify seat belt/material, and determine occupant position. These cameras are also expected to play a vital role in whatever driver impairment (DUI) detection/prevention scheme NHTSA lands on. (Final rulemaking was originally due in November 2024.)

Volvo Multiadaptive Safety Belts

Volvo Multiadaptive Safety Belts leverage info from occupant sensors and myriad others around the car—including those sensing an impending or occurring crash—to tailor the actions of the seat belt retractors and force limiters to each individual occupant, expanding the load-limiter profiles from three to 11. The idea is for the vehicle to know exactly who is sitting where, in what condition, and in what posture before deciding how restraints should behave, thereby providing maximum protection while causing minimal harm.

Suppliers and OEMs are increasingly focusing on that brief window between prediction of an imminent crash and the actual impact (which in some cases can be a second or more) to tailor suspension height, close windows and sunroofs, and make small adjustments to seating position that can improve occupant kinematics. Improvements in side-impact crash prediction could even make ZF LIFETEC’s pre-crash side airbags viable. The company has a traditional seat-mounted one that deploys in two stages, the first being 200ms before impact. It has also shown an external side-impact airbag that deploys from the vehicle’s rocker panel. This 10–14-cubic-foot nylon bag adds about 15 inches of valuable “crush space” during a side impact. Clearly, ensuring against costly false deployments is crucial to the effectiveness of either.