Lidar sensors capture a three-dimensional image of a space. They shoot out invisible (infrared) beams of light and then measure how long it takes for the beam to reflect back. A single lidar sensor will shoot out multiple beams of light and will rotate very rapidly.
By repeatedly scanning an area, it builds up a picture (called a point-cloud) of the space:
Software analyzes those points and looks for clusters of points that move together through the space. It’s not hard to see in the image above how those points can be resolved into objects and then tracked over time. In fact, the point clouds are usually detailed enough that it’s quite easy to distinguish different types of objects. The size and shape of the point cloud make it possible to distinguish cars from people, adults from children, and luggage or carts from people. What’s more, because lidar sensors are creating those point clouds multiple times a second, they can track each object’s persistence over time. That means a lidar sensor (or an array of sensors matrixed together), can track individual objects as they move across an entire location.
Because lidar sensors are generating their own light, they work in any lighting conditions – day or night, indoors or outdoors, in any type of location from a retail store to an airport to a zoo. Each sensor can potentially cover quite a large area (think a football field – but with many practical limitations that limit range), and when lidar sensors overlap the underlying software can seamlessly combine point-clouds from multiple sensors. This lets the sensors maintain tracking of an individual even when one sensor’s line of sight gets blocked by an obstruction or another person, and it lets an array of sensors track objects persistently across very large spaces.
Best of all, lidar sensors CANNOT be used to identify an individual. Even the very best lidar produce images produce images like what you see above. The level of detail isn’t good enough to know who somebody is. Even if a lidar sensor is hacked by a sophisticated intelligence agency or bad actor, it can’t be used to identify a specific person. Lidar can’t be subsumed into a push-marketing system either. It really is JUST a measurement and analytics technology.
This combination of coverage and privacy makes lidar a nearly perfect fit for people-measurement analytics.