Invariant Smoothing

Invariant smoothing is a powerful estimation technique leveraging Lie group theory to address challenges in state estimation, particularly in robotics and navigation. Current research focuses on applying invariant filters and smoothers to diverse problems, including robot localization using IMU and GNSS data, and handling sensor biases and unknown parameters like wheel radius or GNSS antenna position. This approach offers improved robustness and accuracy compared to traditional methods, particularly in challenging scenarios with low process noise or limited sensor data, leading to more reliable and efficient solutions for autonomous systems.