28-10-2020 | By Robin Mitchell
Researchers from the University of Nevada have been experimenting with LiDAR in traffic for the past few years, and now their system is being further established. What is LiDAR, what advantages does it have over cameras, and how can it help reduce traffic issues?
LiDAR is an acronym for Light Detection and Ranging and is a technology near identical to RADAR. LiDAR uses a pulsed beam of light (often a laser), and a rotating head/mirror that quickly scans the area with the pulsed beam of light. On each pulse of light, an internal counter is started, and when the beam of light hits a distant object, it bounces back (the amount of light returned depends on the reflectivity of the object). A sensor detects the reflected beam, and the counter is then stopped. Since the frequency of the counter is known, as well as the speed of light, the distance between the object and the LiDAR sensor can be determined. Other light ranging techniques exist, such as those that utilise phase shifts to produce an accurate measurement, but these are often seldom used in high accuracy laser measurement systems that don’t map.
Images generated by LiDAR have significantly lower resolution than those of standard images, and standard images allow for measurement of colour, reading of text, and detection of faces. However, LiDAR is a distance map as opposed to a standard image; thus, the resulting image is made up of distance measurements. Static images from a regular camera cannot store distance, and distance data from still images requires two cameras (using the stereoscopic effect), which is a complex process. This allows for LiDAR to create a 3D view of the immediate surrounds, and the combination of multiple distance measurements allows for the determination of the speed of detect surfaces.
Researchers from the University of Nevada have spent the past few years experimenting with fixed LiDAR systems to monitor traffic. Back in 2017, Professor Hao Xu set up a roadside LiDAR sensor that would map the surroundings and look for cars, pedestrians and objects. What makes this setup different to standard LiDAR systems is that LiDAR is generally used in self-driving vehicles as opposed to static positions along a road. The data gathered by the LiDAR system was useful, and the team has now been given a boost to integrate more LiDAR systems across various roads.
The data gathered by the LiDAR system allows for the detection of cars, pedestrians, object, and their relative speeds by using specially developed algorithms by the research team. The result is the ability to detect speeding vehicles quickly, plot speed/time graphs showing when people slow down and speed up if pedestrians are on the road, and how cars behave around traffic stops and junctions. According to the research team, this data can further be used to improve traffic systems by applying more efficient traffic control measures at crossings, identify areas that have increased risk of accidents, and improve road design.
The use of fixed LiDAR systems can not only provide historical data on traffic but may be able to provide real-time information. Such a system would most likely operate on the edge, as 3D mapping data would need to be processed in real-time. However, even then an internet connection would still be required, and the biggest candidate to date would be 5G due to its low latency, high speed, and eventual wide-spread coverage. Thus, LiDAR-enabled traffic monitoring would become a key piece of infrastructure for a smart city.