conferences | speakers | series

Loco Positioning: An OpenSource Local Positioning System for robotics

home

Loco Positioning: An OpenSource Local Positioning System for robotics
FOSDEM 2017

Positioning in robotics has alway been a challenge. For outdoor, robots GPS is solving most of the practical problems, but indoor, precise localization is still done using expensive proprietary systems mainly based on an array of cameras.

In this talk, I will present the loco positioning system: an open source Ultra Wide Band radio-based local positioning system, why we need it and how it works. I will also speak about its usage with the Crazyflie 2.0 open source nano quadcopter, of course ending with an autonomous flying demo.

I am Arnaud Taffanel, co-founder of Bitcraze. Bitcraze is designing and making the Crazyflie 2.0 nano quadcopter, a flying open source development platform. We are designing, making and selling hardware platforms and releasing all software as open source projects on Github.

From the beginning of the Crazyflie project, when it was still an after-work fun project, we were dreaming of a Local Positioning System that would allows us to fly the quadcopter autonomously. For years, the only viable solutions where very expensive camera-based motion capture systems. We have been assisting researchers using the Crazyflie with this kind of system but it was out of our reach and out of reach for most of the community.

About a year ago, we discovered that a company, Decawave, had released an ultra-wide-band radio module that would allows us to build an Local Positioning System for a fraction of the price of the motion captures system. This became the Loco Positioning System and it is still under heavy development with the help the community. The Loco Positioning System hardware has been released last summer in "early access" which means that the hardware was finished and tested but that much of the software and algorithm were still under development.

We now have a flying system. With the help of researchers and industrials, we are starting to have state-of-the art algorithm for position estimation and autonomous flight. Currently the main users for the LPS and Crazyflie are universities for research and education, technical artists for shows and industrials for research and tech demos. In the future, technologies and algorithm developed for the LPS will be used for even-more affordable positioning system (for example based on webcams) to allow anyone to play with autonomous flying platforms.

Speakers: Arnaud Taffanel