Currently odometry is generate directly in values of {x, y, theta}. Ideally, it should follow the same model we've been using (by Thrun, Probabilistic Robotics). The following approach will be used
Generate odometry in values of {translation, rotation} in the local frame.
Robot converts the new odometry values to {x, y, theta} using the previous orientation value.
Currently odometry is generate directly in values of
{x, y, theta}
. Ideally, it should follow the same model we've been using (by Thrun, Probabilistic Robotics). The following approach will be usedGenerate odometry in values of
{translation, rotation}
in the local frame.Robot converts the new odometry values to
{x, y, theta}
using the previous orientation value.