Beyond the Algorithm Challenge

Anote’s solution is a human-centered, rapid flood-mapping system that fuses neuromorphic, in-memory, and quantum computing to generate near–real-time flood-extent maps at the network edge. Designed to overcome latency and power limitations of traditional pipelines, the system uses a spiking neural network on neuromorphic hardware for ultra-fast, low-power image segmentation; in-memory resistive RAM arrays for high-throughput spectral and texture preprocessing; and quantum kernel methods to refine uncertain flood boundaries. A browser-based interface enables hydrology experts to annotate edge cases, actively retraining the system with minimal supervision. The platform integrates seamlessly with NASA’s FLOODS data stream and supports drone or ground-based deployment, delivering over 90% mean intersection-over-union in lab tests and reducing time-to-map by 5x. With an active-learning loop and hybrid computing stack, Anote’s solution provides accurate, energy-efficient flood mapping tailored for emergency response, infrastructure planning, and climate resilience applications.