Dead reckoning is a fundamental navigation technique that estimates a current position by mathematically projecting forward from a previously known location using measurements of speed, direction, and elapsed time. In modern GNSS-enabled systems, dead reckoning serves as a critical complement to satellite positioning, providing continuous navigation capability during periods when GNSS signals are unavailable, degraded, or unreliable.
The term derives from historical maritime navigation where sailors would determine their position by ‘deduced reckoning’, calculating distance traveled from speed estimates and compass headings since their last confirmed position. Today’s dead reckoning systems apply this same principle using electronic sensors: Inertial Measurement Units (IMUs) containing accelerometers and gyroscopes measure acceleration and rotation rates, while wheel odometry sensors in vehicles track distance traveled based on wheel rotations.
Dead reckoning is particularly valuable in GNSS-challenged environments where satellite signals cannot penetrate or arrive corrupted. Common scenarios include driving through tunnels, navigating multi-level parking structures, traversing urban canyons with tall buildings blocking satellite visibility, and operating in dense forest canopy. During these GNSS outages, dead reckoning maintains position estimates by continuously integrating sensor measurements to track movement relative to the last known position.
However, dead reckoning accumulates errors over time, a phenomenon called drift. Small inaccuracies in acceleration and rotation measurements compound with each calculation cycle, causing the estimated position to gradually diverge from the true location. Without periodic corrections from an absolute positioning source like GNSS, dead reckoning errors can grow to unacceptable levels within minutes or even seconds depending on sensor quality. Modern positioning systems therefore implement sensor fusion algorithms that tightly couple GNSS with inertial and odometry sensors, using GNSS fixes to bound dead reckoning errors while relying on dead reckoning to bridge GNSS gaps and smooth position solutions.
