NUYHI is an on-device, multimodal AI drone system that detects water-quality anomalies from the air and links them to targeted underwater surveys. A detection drone collects seven image bands—panchromatic, blue, green, red, red-edge, near-infrared, and long-wave infrared. Its onboard edge computer registers and analyzes the images within 2.4 seconds, identifies suspected anomaly locations, and generates survey coordinates without sending raw imagery to an external server. A separate sensing drone travels to the selected coordinates either autonomously or after operator approval. It lowers a modular sensor package to depths of up to 30 m to measure 13 water-quality parameters at different depths. A dual-frequency echosounder records bottom topography, sediment layers, and submerged structures, including their clearance from the bottom. Image data, AI results, GPS coordinates, depth measurements, acoustic profiles, and mission records are linked by time, location, and mission ID as one traceable survey record.