Latest From The Blog


Laser Localization

What's the problem? Motivation This is a blog post designed for non-technical people to understand the basics of localization. As such we are not going to touch any code but rather interactively present you with examples how these systems run and what makes them tick. Localization is an essential part of any autonomous vehicle. Nowadays any self-driving car would use a combination of AI-driven tools to understand where it is, given some previously obtained map. But how do you obtain a map? Using sensors, of course - all autonomous cars have multiple cameras and lasers that record the environment around them. Think of these as a collection of pictures as the one on the left, or a collection of points in space (right). These pictures or clouds are stored on the on-board computers of each vehicle and processed.   Okay I got the map, but we have GPS for determining where we are? Yes that's true, but in addition to being largely inaccurate, it's also not available in covered areas. Have you ever tried using Google Maps in Venice or any forest covered with interleaved trees? You'll notice it jumps from one path to the next. Example Imagine you are in the middle of Spain travelling from Barcelona to Madrid, but you don't have any "smart" navigation to guide you. You have bypassed the ambiguous sign "Terraga " and you see 3 plausible right turns. Your friend is sitting next to you [...]

By |November 30th, 2018|Categories: ORI Blog|Comments Off on Laser Localization
Load More Posts