Natalus is named after a genome of the bats species, and just as a bat uses echolocation to spot its prey and visualize a dark night’s landscape, Natalus devices uses lidar sensors to map out the space around it.
“We initially thought to drill holes in every single parking spot, but that’s just so inefficient,” Terry said.
Cameras wouldn’t work either because of the complex algorithms that have to be adjusted for each separate application. Then Terry and the team began researching sensors.
“We bought a one-dimensional lidar device and started playing with that and some virtual simulations,” he said. “I then taught myself how to analyze and create the artificial intelligence for that.”
Lidar (originally an acronym for Light Detection and Ranging) sensors create 3D maps of the environment by shooting out photons of light. The photons bounce off an object and return to the receiver. The device then computes the time of flight of each photon and places a point. What’s left resembles a black-and-white Seurat painting: an assortment of white dots, which, when put together, form a recognizable image.
One of the biggest challenges of building a device like this is ensuring it can capture the entire landscape.
“In a parking facility,” Terry said, “you need the full 360-degree horizontal and 180 degree vertical.”
Lidar is the same technology used by self-driving cars, but the cameras atop those autonomous vehicles don’t need to capture the same dataset as Natalus’ sensors. Driverless cars capture a 360-degree horizontal but only about a 31-degree vertical since, for driving purposes, you don’t need to map the sky.
Natalus’s sensors hang upside-down (another bat-like trait), however, so they need to have the ability to capture everything below them – the cars, the floor, the open spaces – to give a dataset that parking facilities owners and operators can use.