Today's autonomous vehicles rely on cameras, radar and LiDAR sensors to understand their surroundings and avoid obstacles on the road. Boston, Massachusetts-based startup Teradar believes it can beat those technologies at their own game by looking elsewhere on the electromagnetic spectrum.
The company's solid-state sensors use the terahertz frequency band that's found above radar and below infrared (which LiDAR uses) for enhanced vision through challenging weather conditions like rain, snow, and fog – regardless of how bright or dark it is out there.
Teradar says this makes for native resolution that's up to 20x better than any of today’s radar sensing tech, with vision ranging longer than 984 ft (300 m). That'll allow for Level 3 autonomy, meaning the car can reliably drive itself on city streets and highways, but a human driver will need to be at the wheel to take over when needed.
The idea is that its less-explored approach will not only be better than current-gen sensing tech, but also be easy for automakers to adopt at a lower cost.
The company told TechCrunch it's been working on terahertz sensing systems for the last few years, demoing it for car manufacturers, and most recently, landing US$150 million in a Series B funding round.
Its Modular Terahertz Engine is expected to land at a cost of a few hundred dollars, a great deal less than vehicle brands would spend on combined radar and LiDAR systems for a single car. And while Teradar isn't the first team to leverage this band of the electromagnetic spectrum, it might well be the only one to have developed it for automotive applications – and certainly the first to get this far with it.
Some brands like Mercedes-Benz already offer Level 3 autonomy in their higher-end cars where permitted, and others like Chrysler are getting close to bringing this to market. If Terahertz can indeed upend existing self-driving tech at a lower price point, as it claims, we could start to see more automakers adopt these systems for a wider range of vehicles.
It might still take a little while, though. As Teradar noted, convincing companies to install its sensors and put them through their paces takes time, as does homologation, integrating with services like Google Maps and clearing regulatory hurdles before we can see these on more roads.
Sources: TechCrunch, Teradar
