Edge Vision AI

SOM-1200 Powers Taiwan’s “AIoUT” Smart Sewer Solution

Building an Intelligent Underground IoT System with MediaTek Genio 1200 + Gateway Technology

Taiwan’s urban areas are transforming their underground infrastructure through “The Artificial Intelligence of Underground Things (AIoUT)”, a next-generation smart sewer solution that integrates AI inference and IoT technologies.
This system leverages MediaTek’s Genio 1200 platform and AMobile’s high-performance SBC (Single Board Computer) to bring intelligence, connectivity, and sustainability into rainwater sewers, wastewater pipelines, and potable water networks.

Reimagining the Underground Network with AI + IoT

Unlike surface-level IoT devices, underground environments pose significant challenges — signal attenuation, battery constraints, and maintenance difficulty.

The AIoUT architecture installs advanced sensors in storm drains, sewage systems, and water pipelines to detect flow rate, obstruction, leakage, or contamination. Data is transmitted via long-range Wi-Fi HaLow and processed at the edge using Genio 1200’s AI capability, allowing cities to:

  • Detect sewer blockages and illegal industrial wastewater discharge
  • Identify stormwater and sewage mixing events
  • Monitor underground water levels and prevent urban flooding
  • Detect potable water leaks that may cause road collapse or safety hazards

Why MediaTek Genio 1200 Is the Core of the System

MediaTek’s Genio 1200 (6 nm) chip integrates an octa-core CPU, Mali-G57 GPU, and a 4.8 TOPS NPU, offering industry-leading performance-per-watt for Edge AI applications.

AMobile’s SBC built on Genio 1200 delivers:

  • Superior compute power — surpassing the combined CPU/GPU performance of Raspberry Pi V and Jetson Nano
  • Lower total system cost — approximately half of that traditional dual-board setup
  • AI video analytics — enabling real-time inference for sewer monitoring, water-level detection, and worker-safety monitoring
  • Flexible OS support — Android, Linux Yocto, and Ubuntu
  • Extended temperature range (-20 °C ~ 70 °C) for industrial deployments
  • This cost-effective platform makes edge inference feasible in harsh underground conditions while reducing cloud dependency and network latency.

From Flood Prevention to Water Resource Management

The AIoUT project implements AI inference on rain-sewage mixing analysis using Genio 1200 SBCs to save hardware cost while improving responsiveness.
In New Taipei City, for example, the system monitors 180 × 180 cm drainage channels; when the water level exceeds 115 cm, sensors trigger alerts to pumping or pressurizing stations and send rainfall data to the Central Weather Bureau for big-data analysis and early-warning decisions.

Edge-AI Applications Enabled by Genio 1200

  • AI-based construction safety monitoring: Uses real-time video analytics to identify unsafe worker behavior and environmental anomalies, enhancing safety in underground operations.
  • AI-driven rain-sewage flow intensity analysis: Analyzes real-time water level, flow, and quality data to assess rain–sewage mixing for better drainage planning.
  • AI water-pipe leakage detection: Detects leaks using pressure changes, vibration data, and AI models to reduce losses and prevent ground collapse.
  • AI flood prediction and alert system: Uses rainfall, water-level data, flood history, and SWMM modeling to predict urban flooding and issue early warnings.

Empowering Smart City Sustainability

Through collaboration between MediaTek’s advanced Genio 1200 SoC and AMobile’s edge computing expertise, Taiwan’s AIoUT initiative demonstrates how AI + IoT can safeguard cities from underground risks, reduce environmental pollution, and build resilient, sustainable infrastructure for the future.

Go Back