New LiDAR Sensors showing us the future of ubiquitous 3D Scanning

Quanergy's Mini Solid-State LiDAR has just been unveiled to the market, and it is a peek into the future.

For those of you who don't know what LiDAR is, it is a technology for using lasers to scan a surface and create a 3-dimensional model. It's much more precise than stereoscopic methods, and until recently only used in very specific industrial and military applications. That's in part because until recently the only LiDAR sensors available were large and expensive.

A scanner in your pocket, in your car, or on your drone

But that is changing. Velodyne's Puck-Lite came out earlier this year weighing around 500g and putting LiDAR well within the operating parameters of most commercial drones. Quanergy's new sensor is solid state, weighs in at 100g, is also aimed at drones, and should be available some time next year. (Velodyne,Continental, and others also offer solid-state LiDAR aimed at the autonomous vehicle market). Terabee even has a teeny-tiny LiDAR suitable for home robotics projects.

Clearly, where this is heading is ever lighter and ever smaller LiDAR sensors that will increasingly be found on commercial drones and self-driving cars. But now that the size and weight of these things is already comparable to an iPad Mini -- and shrinking -- it won't be long before we begin to see them used in all kinds of 3D scanning applications, including robotics and even hand-held devices.

LiDAR enables better AI

At Reality Analytics, we're particularly excited about widespread adoption of LiDAR technology. In the drone space our AI tools are already working with LiDAR signals combined with images collected from commercial UAVs to automate a variety of mapping and surveying tasks, identifying surfaces and spotting surface discontinuities. ("That flat spot over there is poured concrete, and there's a crack in it." Or, "on this job site we have 7 different materials stockpiles, which collectively contain 5000 cu meters of sand and 1200 cu meters of gravel." Or, "this part of the crops have been damaged by disease." Or, "is there any deep rust or corrosion on this equipment.") We can do it without LiDAR -- photogrammetry remains a perfectly valid way to generate 3D surfaces from imagery. But LiDAR gives very rich texture information that our AI algorithms can exploit to good effect.

3D scanning everywhere... coming soon

Just wait until we can get that same kind of rich texture data from other surfaces scanned by LiDAR. Buildings and terrain scanned by drones... Roads and roadsides scanned by autonomous vehicles... Indoor and outdoor environments scanned by robots trying to navigate their surroundings.... Scans of rooms and objects by people using portable devices... Much of the world around us will be scanned and available for processing and interpretation. Every device we use will be aware of the 3D world around us - shapes, surfaces, and textures ; what its made from, whether its slippery, whether there are flaws. So much more to come!

#smartcar #artificialintelligence #machinelearning #3dscanning #LiDAR