Human navigation strategies and their errors result from dynamic interactions of spatial uncertainties

Area of Science:

• Cognitive Science
• Neuroscience
• Robotics

Background:

• Human navigation integrates self-motion and landmark cues under uncertainty.
• Current models often treat distinct navigational strategies separately.

Purpose of the Study:

• To provide a unified computational model for goal-directed navigation.
• To explain the integration of spatial uncertainties in perception, cognition, and action.

Main Methods:

• Developed a computational model of probabilistic path planning using optimal feedback control under uncertainty.
• Analyzed how sequential egocentric landmark observations form an allocentric cognitive map.

Main Results:

• The model unifies diverse human navigational strategies.
• It quantitatively predicts navigation errors and variability across experiments.
• It explains the use of internal maps for planning and movement execution.

Conclusions:

• A single framework explains complex human navigation behaviors.
• The model reconciles seemingly contradictory findings in cue-integration research.
• Continuous dynamic interactions of spatial uncertainties are key to navigation.