Massachusetts Institute of Technology (MIT) engineers have developed a new lidar chip that offers a wider and clearer view without moving parts. This advancement could lead to smaller, more durable sensors for self-driving cars, drones, and industrial applications. The technology addresses limitations in current lidar systems, which are often large, costly, and prone to wear.
The new silicon-photonics chip controls light instead of electrical signals. Existing chip-based lidar systems typically have a narrow field of view. They struggle to scan areas at the edges of a scene. Previous attempts to expand this range often increased noise and reduced accuracy.
The MIT team created an array of integrated antennas to solve these issues. This design significantly limits unwanted crosstalk, which occurs when neighboring antennas interfere with each other. The chip can scan a broader field of view with less noise than other silicon-photonics-based methods.
Traditional lidar systems use a rotating unit to direct light pulses. Silicon-photonics-based lidar uses an integrated optical phased array (OPA) to scan a light beam electronically. The OPA uses a group of integrated antennas. Each antenna has tiny, regularly spaced corrugations that scatter light upward and out of the chip.
Researchers control the outgoing beam's direction by changing the phase of light sent to each antenna. This allows the beam to be steered without moving physical components. The MIT team designed antennas with distinct shapes, varying in width and corrugation patterns. This reduces coupling between closely spaced antennas. The design lowered coupling from approximately 100 percent to about one percent in experiments.
The system accurately steered the beam across a broad field of view. It did so without producing grating lobes, which are unwanted beam copies that can confuse sensors. This combination of wide scanning, low interference, and strong beam quality addresses a major challenge in integrated lidar technology. The researchers plan to refine the method to cover an even broader viewing range.
Related stories
Tesla launches Cybercab robotaxi service in Austin with driverless vehicles
Elon Musk's electric vehicle and clean energy company, Tesla, has officially launched its highly anticipated Cybercab robotaxi service in Austin, Texas. This ne...
Druckenmiller used AI to draft critical Wall Street Journal opinion piece
Billionaire investor Stanley Druckenmiller recently disclosed his use of artificial intelligence in drafting an opinion piece highly critical of Scott Bessent. ...
Google acquires anonymized Spirit Airlines data for AI training amid union concerns
Google's acquisition of data from the now-bankrupt Spirit Airlines for a reported ten million dollars has ignited a contentious debate surrounding data privacy ...