in ,

Silicon Chip Improves LiDAR Systems Without the Use of Electronics

Silicon Chip Improves LiDAR Systems Without the Use of Electronics

A silicon chip with no moving parts enables compact, efficient lidar implementations.

Researchers at the University of Colorado describe a new silicon chip that has the potential to move beyond the bulky, complex Light Detection and Ranging (lidar) systems that are hindering the systems’ widespread adaptation.

Light Detection and Ranging (lidar) systems are based on electronically measuring the time it takes a laser pulse to reach its target, and then return, thereby establishing its distance. It will be a fundamental component of any self-driving car.

Google self-driving car.
Google self-driving car with preset day lidar system. Image used courtesy of the University of Colorado

However, existing systems are too expensive and bulky for automotive applications, let alone smartphones, tablets or video games. They employ large, physically rotating mirrors to steer the laser beam about an area of which a “picture” is desired. Researchers at the University of Colorado intend to change that.

“We’re looking to ideally replace big, bulky, heavy lidar systems with just this flat, little chip,” said Nathan Dostart, lead author of a study published in Optica. The study describes a silicon chip containing no moving parts or electronics that will prove practical for real-world applications.

Optical Phases Arrays (OPA) integrated into CMOS platforms, forming small-sized, inexpensive electronic–photonic systems-on-chip that require little power may well turn out to be the linchpin around which a new generation of photonic sensing technologies evolve.

What Are Optical Phased Arrays?

OPAs serve to control the amplitude and phase of light waves transmitted from a two-dimensional surface. By directing the optical properties from the surface, the device controls the direction of the light beam emitted, and it does so with no moving mechanical parts.

Optical phased arrays can form and electronically steer optical beams, such as laser pulses, to be emitted from an on-chip aperture. They can also optimize the reception of the reflected light, completing the travel circuit.

In critical automotive applications such as advanced driver assistance systems (ADAS), this will allow for calculating the distance to a pedestrian, another car, or any road hazard.

Silicon chip OPA
Silicon chip OPA. Image credited to University of Colorado

Wavelength Steering

Dostart and his colleagues have developed a method of steering laser beams called wavelength steering. The key to the technique is pointing each wavelength of the laser at its own unique angle.

Rather than doing this along one dimension, they’ve developed a way to work in two dimensions simultaneously, instead of only one. They’ve employed multiple wavelength patterns to build actual three-dimensional images.

The beams are readily controlled by changing colors. This allows multiple phased arrays to be simultaneously controlled, allowing the creation of a bigger aperture for a higher resolution image. There is an 8 x 4 array of 32 tiles in the picture below, with a slightly different grating design. Illustrated are two matching pairs lit up by the laser.

Their beams superimpose on each other. The far-field beam interference pattern demonstrates the formation of the tiled beam. The next image illustrates a raster scan pattern addressed by the wavelength-steered silicon photonic OPA.

Raster scan pattern.

The curving pattern, reminiscent of a fingerprint, indicates the dispersion affected by the new waveguide system built on a silicon chip.

According to Miloš Popović, associate professor of engineering at Boston University and study co-author, “Electrical communication is at its absolute limit. Optics has to come into play, so all these big players are committed to making the silicon photonics technology industrially viable.”

Source: www.allaboutcircuits.com

What do you think?

486 points
Upvote Downvote

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Rimac C_Two Will Teach You How To Be A Better Driver Using AI

Rimac C_Two Will Teach You How To Be A Better Driver Using AI

Helm.ai applies ‘Deep Teaching’ to Level 2 to 4 autonomous vehicles

[video] Helm.ai applies ‘Deep Teaching’ to Level 2 to 4 autonomous vehicles