Precise real-time localization using comma.ai panda

Organization
Autonomous Driving Lab
Abstract
NovAtel GNSS system in our current car works extremely well and delivers <10 cm accuracy even during driving. To achieve this accuracy, it combines:
* multiple satellite systems, e.g. GPS, GLONASS, Beidou, Galileo;
* information from two antennas;
* inertial measurement unit (IMU) relative positioning;
* real-time kinematics (RTK) correction system.

We believe similar level of accuracy could be achieved with comma.ai panda by combining following sources:
* GPS localization,
* real-time kinematics (RTK) correction system,
* vehicle odometry,
* accelerometer (needs comma two),
* visual odometry (needs comma two).

Kalman filter (for example comma.ai rednose library) can be used to combine different sources. Potential cooperation with T├Áravere Observatory (Hendrik Ehrpais).
Graduation Theses defence year
2020-2021
Supervisor
Tambet Matiisen
Spoken language (s)
Estonian, English
Requirements for candidates
Level
Masters
Keywords

Application of contact

 
Name
Tambet Matiisen
Phone
5286457
E-mail
tambet.matiisen@ut.ee
See more
https://docs.google.com/document/d/1Hd1qjDmfCGvnW2bysQAQvfPB5UEwFDcnClMMFcNCt4k/edit?usp=sharing