Bio-inspired cognitive navigation for robots

Zhaotie Hao, Bin Guo, Yasan Ding, Linwei Li, Shiqi Liu, Tingting Gao, Sicong Liu, Zhiwen Yu

Abstract: Animals navigate unfamiliar environments for hours on minimal energy, yet current robotic navigation systems — whether model-based or data-driven — struggle to generalize, operate within tight energy budgets or respond rapidly enough to changing conditions. These shortcomings reflect a fundamental absence: robots lack the integrated cognitive architecture that underpins biological navigation. This Review examines how biological principles can close this gap, showing that cognitive maps, adaptive memory and hierarchical planning translate into robotic architectures capable of flexible, energy-efficient navigation, and evaluates real-world deployment, including low-power neuromorphic implementations. Bridging biology and engineering will require advances in spatial representation, knowledge use and spatial reasoning, together with sustained interdisciplinary collaboration. Animals navigate robustly by combining cognitive maps, adaptive memory and hierarchical planning, but robots lack this integrated cognitive architecture. This Review examines how biological navigation principles can be translated into robotic systems, addressing spatial representation, knowledge use, multi-scale planning, energy-efficient implementation and the open challenges facing real-world deployment.

External IDs:doi:10.1038/s44287-026-00294-7