Satellite-Enabled Edge Intelligence for Climate-Resilient Emergency Mobility: A Survey of On-Orbit Edge Computing

Climate change is increasing the frequency and intensity of extreme weather events, which can rapidly disrupt road capacity, public transit operations, and logistics corridors. Reliable emergency mobility and low-carbon recovery therefore require timely, trustworthy, and spatially comprehensive situational awareness as well as resilient decision-making under intermittent connectivity. Earth-observation satellites and emerging LEO constellations provide global coverage for hazard monitoring (e.g., floods, wildfires, landslides) and infrastructure assessment, yet the resulting data volumes and limited contact windows make purely ground-centric processing increasingly insufficient for real-time transportation response. This paper surveys on-orbit edge computing (OEC) as a spaceborne edge layer that can close the sensing–analysis–action loop for climate-resilient transportation systems. We organize the literature into three evolutionary stages: 1) single-satellite on-orbit processing for event triggering and data triage, 2) satellite–ground collaboration for split inference, task offloading, and end-to-end orchestration, and 3) inter-satellite distributed computing for constellation-scale learning and inference under dynamic topology. Beyond summarizing methods, datasets, and platforms, we discuss how these OEC capabilities can be integrated with transportation digital twins and emergency operations to enable faster disruption detection, evacuation planning support, and bandwidth- and carbon-aware information delivery. Finally, we outline open challenges and future directions, including robust benchmarking under climate stressors, trustworthy autonomy, and LLM-assisted decision support across satellite–ground workflows.