An 'observer' in navigation algorithm context is

• A. A hardware device measuring external signals.
• B. A state estimator like a Kalman filter that observes sensor data to estimate states.
• C. A data logger with no estimation capability.
• D. A supervisor program that only logs results.

Answer: (B)

An observer is a state estimator that reconstructs the full set of system states from sensor measurements and the system model. In navigation, you often can’t measure every state directly or with perfect accuracy, so you use a model of how the system evolves and combine it with sensor data to estimate hidden quantities like true position, velocity, attitude, or sensor biases. A Kalman filter is a classic example of an observer: it predicts the next state using the motion model and then updates that prediction with new measurements, giving refined state estimates and their uncertainty. This fusion of model and data makes the estimates robust to noise and errors.

The other options describe devices or logging tools that collect data or oversee processes but do not produce state estimates from a model and measurements, so they don’t fulfill the role of an observer.