Inertial Odometry

Inertial odometry (IO) aims to estimate the position and orientation of a moving platform using only inertial measurement unit (IMU) data, crucial for applications like robotics and autonomous navigation. Current research emphasizes improving IO accuracy and robustness by incorporating diverse sensor modalities (e.g., LiDAR, cameras, radar) through tightly-coupled filtering techniques (like Kalman filters and their variants) or data-driven approaches using deep learning (e.g., neural networks with attention mechanisms). These advancements address challenges like sensor noise, drift accumulation, and environmental complexities, leading to more reliable and efficient localization in various scenarios, including dynamic environments and resource-constrained platforms.