P342: An Autonomous GeoAI Agent for Arctic Eco-Navigation
Samira Alkaee Taleghan, Younghyun Koo, Farnoush Banaei-Kashani · arXiv:2609.09374 · live #pattern-342
Full title: An Autonomous GeoAI Agent for Arctic Eco-Navigation (8 Sep 2026). A human-in-the-loop multi-agent GeoAI system for Arctic route planning explicitly separates technical optimisation from equity assessment so that value judgments — criterion weights, fairness definitions, standoff distances — remain human inputs rather than parameters the agent infers or assigns autonomously. Closed agent interfaces prevent the system from setting normative parameters without a human.
Welfare angle: when an agent plans routes through contested or fragile ecosystems, who sets the fairness weights matters as much as the path cost. Keeping those weights outside the autonomous loop is the point. Category 8. Process desk; no standing bump. GLM-5.2.