AI-powered urban issue detection, triage, and dispatch — from photo to fix in seconds.
Photos classified by Llama 4 Scout vision model into 7 categories with severity scoring and chain-of-thought reasoning
Bayesian Poisson-Gamma priors on 20 IoT sensors detect neighborhood anomalies; z-score alerting via SSE
K-Means clustering + 2-opt TSP optimization routes repair crews to minimize travel distance
Graph diffusion models predict issue spread; cross-correlation analysis reveals causal cascades
| Algorithm | Family | Complexity |
|---|---|---|
| Vision LLM Classification | AI/LLM | O(1) per API call |
| DBSCAN Spatial Clustering | Clustering | O(n log n) |
| K-Means Crew Assignment | Clustering | O(n·k·i) |
| 2-opt Route Improvement | Optimization | O(n²) per iteration |
| Bayesian Poisson-Gamma Anomaly | Statistics | O(1) per update |
| Linear Programming Budget | Optimization | O(n·m) simplex |
| Graph Diffusion Prediction | Prediction | O(n³) |
| Cross-Correlation Causality | Causality | O(n·p) per pair |
| Weibull Survival (SLA) | Survival | O(1) per report |
| Perceptual Hashing (pHash) | Deduplication | O(1) per comparison |