Historically, OxTS has focused on two distinct areas of expertise. On the one hand, there’s the automotive side of our business; on the other surveying and mapping. However, the two are being brought ever closer together thanks to the growing proliferation and prominence of one particular sensor technology: LiDAR.
The world around us is becoming increasingly digitised, and LiDAR is one of the sensors at the heart of that. LiDAR – for ‘light detection and ranging’ – creates highly accurate, high-resolution 3D point clouds of the environment and is a well-established technology in the surveying and mapping sector. Its use has become more widespread in recent years thanks to the growing use of drones and unmanned aerial vehicles (UAVs).
Autonomy with LiDAR
It is only more recently that LiDAR has come to prominence in the automotive world, as a sensor component in advanced driver assistance systems (ADAS) and, most notably, on autonomous cars. LiDAR sensors are for now either too large, too expensive or too limited in effective range to become widely used as part of the ADAS functions on production cars. Autonomous vehicle developers, on the other hand, have fully embraced the technology.
An autonomous car needs to analyse a huge amount of data in order to navigate in safety. Much of that data is provided in real-time by a combination of sensors, including radar-, camera-, ultrasonic- and LiDAR-based systems. Together they enable the vehicle to interpret and react appropriately to the environment, other vehicles, vulnerable road users and pedestrians.
LiDAR – mapping in 3D
In order to do this as efficiently and effectively as possible, having access to a pre-programmed, high resolution, 3D map of its surroundings is ideal. In effect, it cross-references what its real-time sensors can ‘see’ with what it already expects to find. This reduces the need for complex real-time analysis of the infrastructure-based surroundings which, without the 3D map, would otherwise be an entirely alien and unfamiliar environment.
Point cloud generated with OxTS INS and Velodyne LiDAR.
The creation of these hyper-detailed 3D map datasets has become a growth industry in itself. Growing numbers of autonomous vehicle developers and tech startups working to to create hyper-detailed 3D maps for self-driving cars to use as a continuous reference or to aid real-time performance development. But while LiDAR can be used to create these maps, what it can’t do is tell you precisely where in the world that data relates to.
Frame of reference
“LiDAR has no concept of where it is at any given time,” says Paris Austin, OxTS’ business manager for LiDAR survey applications. “A GNSS/INS allows you to georeference your LiDAR point cloud (also known as point cloud localisation) both in absolute (Earth frame) and relative (XYZ) location, and it can even do it in real-time if you need to.”
“If you’re looking for real-time obstructions around the car, you don’t necessarily care where you are in the world, only what’s in a 100-metre or so radius of you. But without the GNSS/INS you can’t hope to cross-reference that against where the vehicle thinks it is, either within the lane or in relation to, say, an upcoming zebra crossing, a ‘stop’ sign or other infrastructure-based features. And that’s what the 3D map they created prior to the real-time test will help with.”
“At every stage in that process we’re involved in some way. Time synchronisation of sensors is a major task. GNSS is ideal for that and if you want to know where you are on the Earth to within a couple of centimetres and at a high data rate, a GNSS/INS is one of the easiest and most readily available ways of achieving it. For both the real-time applications and the initial 3D map creation, there’s a place for what we’re doing in terms of removing unknowns, because we give you that high-accuracy time, position and orientation data at every step along the way.”
Know where on Earth
LiDAR as a technology will be increasingly adopted to create 3D maps of environments, not just by the automotive industry, but also in surveying, agriculture, construction infrastructure and more. However, it is the marriage LiDAR with GNSS/INS technology that tells you the location and orientation of that environment in relation to the Earth. Without this, the LiDAR is often worthless.
Visit our page on efficient and reliable georeferencing solutions to learn more.