Urban Operations Scheduling and Decision Support

Public-safe evidence only: figures are selected from deployment and model outputs without sensitive operations records.

System Context (One Plant + Two Stations)

The deployment focuses on a one-plant-two-stations urban service scenario where scheduling decisions must balance service quality, energy cost, and operational stability.

Demonstration area map for one-plant-two-stations water dispatch

Demonstration-area map for South City Water Plant with Fuxing and Renmin Square pumping stations.

中文图注：黄浦示范区“一厂两站”供水区域示意图。

Model & Scheduling Framework

Water dispatch method framework from multi-source sensing to scheduling generation

Framework from multi-source sensing and event triggers to robust dispatch-plan generation.

中文图注：多源感知-事件触发-调度方案生成的方法框架。

Forecasting and scheduling workflow connecting demand and pressure estimation with operational decisions.
Constraint-aware scheduling with operator-configurable rule sets.
Deployment-ready integration with dashboard-based review and fallback strategy options.

Forecast Performance Evidence

Pressure forecasting example with MAPE shown on figure

Pressure forecasting example with low MAPE under full-period operational data.

中文图注：压力全时段预测示例图（图中包含 MAPE 指标）。

Dispatch Platform in Operation

Production dashboard for real-time scheduling status, service map, and execution tracking.

中文图注：供水智能调度主界面，包含压力热力分布与执行日志。

Dispatch scoring and adoption evaluation panel

Dispatch-scoring panel supporting operation-quality assessment and strategy adoption review.

中文图注：调度评分界面，用于评估指令准确性、压力达标率与能耗表现。

Measured Impact

Metric	Reported change
Peak energy consumption	~10% reduction
Pump switching frequency	~20% reduction
Pressure compliance	~3% improvement
Operational mode	7x24 continuous dispatch

Improved service-level stability with lower actuation burden.
Better operator interpretability through dashboard and scoring modules.
Reusable strategy templates for recurring operating patterns.