New Silicon Chip Development Could Revolutionize LiDAR
Jacob Bourne posted on July 16, 2020 |
LiDAR advancement to propel automation and augmented reality technologies.
Silicon chip with 32 serpentine optical phased array (SOPA) tiles for use in small-scale LiDAR system. (Image courtesy of Bohan Zhang and Nathan Dostart.)
Silicon chip with 32 serpentine optical phased array (SOPA) tiles for use in small-scale LiDAR system. (Image courtesy of Bohan Zhang and Nathan Dostart.)

Light detection and ranging (LiDAR) is a technique that utilizes laser beams to measure distances in an environment. Like an optical version of echolocation, the laser beams are bounced from objects to a sensor that maps a 3D image of the surroundings in real time.

Now, new research seeks to make LiDAR smaller and cheaper. A research team at the University of Colorado (UC) Boulder recently published a study that discusses a new form of LiDAR that allows the technology to avoid the large rotating mirrors that currently steer laser beams to create a 3D image.

The new approach uses what the researchers call a serpentine optical phased array (SOPA), which serially connects an array of low-loss grating waveguides to provide passive beam steering. The researchers claim that this system is space efficient and scalable to long-range LiDAR systems, improving the resolution and scanning speed needed to map a complex environment.

“We’re looking to ideally replace big, bulky, heavy lidar systems with just this flat, little chip,” said lead researcher Nathan Dostart.

The SOPA approach steers laser beams by wavelength, with different wavelengths corresponding to different angles. It works along two dimensions simultaneously, and multiple phased arrays can be used together for a larger aperture and higher resolution image. The researchers experimentally demonstrated “the first SOPA using a 1450–1650 nm wavelength sweep to produce 16,500 addressable spots in a 27x610 array.”

A Google self-driving car from 2016. Note the bulky LiDAR hat. (Image courtesy of Wikipedia user Grendelkhan.)
A Google self-driving car from 2016. Note the bulky LiDAR hat. (Image courtesy of Wikipedia user Grendelkhan.)

LiDAR is notably used in autonomous vehicles to allow cars to navigate complex roadways effectively and without collisions (though collisions aren’t always avoided). The systems are currently quite large, as evidenced by the cumbersome-looking apparatuses on the roofs of some Google self-driving cars. LiDAR is also used as a scientific instrument to map geographic features such as shorelines and elevation, as well as in a variety of applications ranging from forestry, agriculture, space missions, emergency response operations and video games. Apple’s latest iPad Pro even incorporates LiDAR.

This advancement in LiDAR could solve a key problem for the development of autonomous vehicles that depend on it to fully replace human drivers. To date, LiDAR systems represent the most expensive component of self-driving cars, with some systems costing up to $70,000 each. Such costs are prohibitive to expanding the use of the technology in everyday consumer devices. Studies such as the one by the UC Boulder team are part of a push to harness silicon chips to make LiDAR simpler, smaller and less expensive, and thus much more easily used in self-driving cars, smartphones or video games, for example.

“Electrical communication is at its absolute limit. Optics has to come into play and that’s why all these big players are committed to making the silicon photonics technology industrially viable,” said coauthor Miloš Popović. 

This innovation is already looming on the horizon with Apple rumored to be planning to include a LiDAR camera in its upcoming iPhone 12. It’s hoped that advancements in LiDAR for consumer devices will aid in facial recognition security, assist with mapping out hand and footholds for climbing routes and identify wildlife, along with countless other applications.

“We’re proposing a scalable approach to LiDAR using chip technology. And this is the first step, the first building block of that approach,” said Dostart. “There’s still a long way to go.”

Recommended For You