Las Vegas using cameras, LiDAR to make downtown safer

Las Vegas is now deploying infrared cameras and LiDAR-based sensors to track vehicle and pedestrian movement in its busy downtown area to improve safety, according to a recent article on the online site of StateScoop.

According to the article, the city began a year-long pilot project in late 2018 with the Nevada state government, Dell and Japanese telecom company NTT to use Internet-connected audio and video sensors to manage vehicular, pedestrian and bicycle traffic. This attempt focuses on making downtown Las Vegas’ one-way streets safer, said Michael Sherwood, Las Vegas‘ director of innovation and technology.

“If we knew how many accidents that were almost occurring as well as ones that did, would we change our methodology on installing traffic-calming measures or pedestrian measures to help ensure the proper signage is there?,” Sherwood was quoted as saying.

According to Sherwood, NTT’s optical sensors, placed in intersections or on light poles around the city, can detect near misses as well using infrared vision or LiDAR. “What we’ve found out is while we didn’t have a lot of accidents on a [one-way] street, we did have a lot of people going the wrong way,” he said.

RELATED: LiDAR system promises major benefits for autonomous vehicles

RELATED: LiDAR System Raises Safety-Performance Bars

The sensors also collect audio information to help discern vehicle location and use Dell’s software to compute the audio-visual data on the edge, or within the sensor itself. Metadata is then sent back to a central repository for the city to analyze.

“Instead of sending all the data back to a core, trying to analyze it and send something back, even though that might take milliseconds, it really is not helpful if you’re trying to change a light from green to red based on a condition,” Sherwood was reported as saying.

Sherwood noted that the infrared and LiDAR technologies provide the all-important task of detecting the shape of an object, rather than providing fine details about an object.

“We just need to know that a vehicle went the wrong way,” he said.