Orientation and Offsets

In order to produce accurate data, the rover needs to know the position and orientation of all attached sensors. These sensors include the IMU as well as a combination of one or multiple GNSS antennas, LiDAR sensors and cameras.

In this section, the terms offset and translation will be used as synonyms. However, the terms rotation and orientation are not the same; a sequence of rotations around different axes leads to an orientation. For every orientation, there is an infinite number of sets of rotations that leads to that orientation. Generally, a sensor will be mounted, translated, and then rotated relative to the IMU. When the terms “translation” and “rotation” are used together, they form a “transform”.

The system’s frame of reference is the IMU (which is why the reference point in the IMU is named the “center of navigation”).

The offsets for all sensors are specified as offsets from the IMU to the sensor, not the other way around (sensor to IMU). However, the axis definitions used differ by sensor type. The rotations for all sensors are specified using different conventions depending on the sensor type.

When mounting the rover to a vehicle, it is important to have an idea of the vehicle’s forward and up directions. While some vehicles, e.g. multirotors, don’t enforce a forward-direction, navigation systems do, so even a multirotor needs to be flown like a fixed-wing aircraft (aside from takeoff and landing).