What Is Dead Reckoning?

Dead reckoning is a navigation method in which a moving object's current position is estimated by advancing a known prior position using measurements of speed, heading, and elapsed time, without reference to external landmarks or signals. The technique predates electronics by centuries, originating in maritime navigation where sailors calculated position from logged speed and compass bearing when celestial sightings were unavailable. In contemporary engineering, dead reckoning is implemented in inertial navigation systems (INS) that integrate data from accelerometers and gyroscopes to compute position and orientation continuously. It remains relevant wherever satellite navigation signals are unreliable or unavailable, including underground facilities, tunnels, urban canyons, and contested electromagnetic environments.

The approach draws on mechanics, stochastic estimation, and signal processing. Its central challenge is error accumulation: small biases and noise in sensor measurements are integrated over time, causing position estimates to drift away from the true trajectory at a rate that grows with time since the last external fix.

Inertial Sensors and Error Accumulation

An inertial navigation system performing dead reckoning uses accelerometers to measure specific force along each body axis and gyroscopes to measure angular rate. Integration of accelerometer data yields velocity; a second integration yields position. Integration of gyroscope data yields attitude, which is needed to resolve accelerometer readings from the body frame into the navigation frame before integration. Every integration step propagates and amplifies errors: a constant accelerometer bias of 1 mg produces a position error that grows quadratically with time, reaching tens of meters after only a few minutes of unaided operation. Gyroscope drift compounds this by causing the attitude estimate to rotate, misaligning the acceleration integration frame. Microelectromechanical systems (MEMS) inertial sensors, used in smartphones and automotive applications, have higher noise and bias than tactical-grade fiber-optic or ring-laser gyroscopes but are dramatically smaller and lower in cost. The characteristics of MEMS inertial sensors relevant to dead reckoning are described in depth in NIST guidance on inertial sensor calibration.

Sensor Fusion and Kalman Filtering

Bounding the drift inherent in dead reckoning requires periodic corrections from external observations, which are blended with the inertial estimate through an optimal filter. The Kalman filter is the standard algorithm for this purpose: it maintains a statistical model of the system state and its uncertainty, and when a measurement from an external sensor arrives, it updates the estimate in proportion to the relative trustworthiness of the prediction and the measurement. Extended Kalman filters (EKF) and unscented Kalman filters (UKF) handle the nonlinear dynamics typical of full six-degrees-of-freedom navigation. Complementary sensors include wheel odometers, barometric altimeters, magnetometers, and vision-based odometry. Automotive dead reckoning systems fuse GNSS position fixes with wheel speed sensors and steering angle to maintain continuous position estimates through signal outages, as described in technical documentation from u-blox on automotive dead reckoning technology.

GNSS-Denied Navigation

Dead reckoning is of particular engineering significance in environments where Global Navigation Satellite System (GNSS) signals are unavailable or deliberately jammed. Submarines traditionally relied on pure inertial dead reckoning, accepting growing uncertainty until a surfacing fix could be obtained. Autonomous ground vehicles navigating in tunnels or urban canyons use tightly coupled INS/GNSS architectures that seamlessly hand off between satellite-aided and inertial-only modes. Indoor pedestrian navigation research, reviewed extensively in IEEE Transactions on Instrumentation and Measurement, addresses foot-mounted MEMS IMUs that exploit zero-velocity update (ZUPT) corrections each time a foot contacts the ground, resetting accumulated velocity error and limiting position drift to a few percent of total distance traveled.

Applications

Dead reckoning has applications in a wide range of fields, including:

• Autonomous vehicle localization in GPS-denied or signal-blocked environments
• Submarine and underwater vehicle navigation
• Indoor pedestrian tracking for emergency responder positioning
• Unmanned aerial vehicle navigation during GNSS outages
• Spacecraft attitude and orbit determination
• Robotics and mobile platform localization