Self supervised RF odometry estimates robot motion in tough conditions
Self-Supervised Doppler-Guided RF Odometry
RoboticsMachine Learning
Summary
Robots sometimes need to figure out where they are even when vision is blocked by darkness or smoke. This paper shows how radio waves alone can help a robot track its movement without needing extra sensors or labeled data. The authors created a method that teaches the robot to match radio signals over time by checking if its guesses make sense in terms of motion speed and direction. Their approach works well even when other objects are moving around, making it good for real-world environments.
What this means in practice
- •For robotics engineers: Estimate robot position using only RF sensors in low visibility or cluttered environments without GPS or inertial data.
- •For security system designers: Track moving objects and their own position indoors using radio frequency signals even when visual cameras fail.
Authors
Kunzhe Song, Jialuo Du, Jingkai Lin, Huacheng Zeng
Abstract
Radio-frequency (RF) sensing enables robot ego-motion estimation in darkness, smoke, and fog, yet RF odometry remains challenging because multipath and moving reflectors make cross-frame matching highly ambiguous. Existing methods often rely on auxiliary inertial sensors or ground-truth pose supervision. We present RFWalk, a self-supervised, RF-only odometry framework that models dense correspondences across range-azimuth frames as random walks on a space-time graph and learns them from unlabeled sequences through temporal cycle consistency. To rule out correspondences that are cycle-consistent but physically implausible, we introduce a Doppler consistency loss that constrains correspondence-induced motion using radial velocity measured by the RF sensor itself. Ego-motion is then recovered through robust $\mathrm{SE}(2)$ registration that suppresses uncertain correspondences and those arising from independently moving objects. Without pose labels or auxiliary sensors, RFWalk remains competitive with supervised baselines in static scenes and reduces absolute trajectory error by 58% in dynamic scenes with moving pedestrians.