Inertial measurement unit

An IMU is a compact sensor package that reports how a device is moving. At its core are accelerometers, which measure linear acceleration, and gyroscopes, which measure angular velocity, and many IMUs add a magnetometer for heading. Together these give a high-rate stream of motion and orientation data that other systems build on.

IMUs update quickly and work anywhere, but their measurements drift when integrated over time, so they are rarely used alone.

Accelerometers and gyroscopes sample motion many times per second. Integrating acceleration gives velocity and position, and integrating angular velocity gives orientation, but noise and bias make these integrals drift. Sensor fusion combines the IMU with slower, drift-free references like cameras or GPS to get an estimate that is both fast and stable, and calibration corrects systematic sensor errors.

IMUs are foundational to motion tracking in physical AI, feeding odometry, SLAM, and stabilization. Their high update rate complements slower perception sensors, which is why fusing an IMU with cameras or LiDAR is a standard pattern in robotics.

Linear acceleration from accelerometers and angular velocity from gyroscopes, and often magnetic heading from a magnetometer.

Integrating its noisy measurements causes drift, so it is fused with drift-free references like vision or GPS for reliable positioning.

Odometry, SLAM (simultaneous localization and mapping), Sensor fusion, Sensor calibration.