Back to Projects

Smart Water Usage Analytics – Town of Cary

A comprehensive data analytics project using SAS to manage over half a billion smart water meter readings, enabling early leak detection, customer awareness, and improved municipal water efficiency.

2025
60,000 Smart Meters
0.5 Billion Data Points
SAS Cloud Platform
SAS Analytics Power BI AWS Data Storage Data Visualization

Project Overview

Objective

To analyze massive real-time water meter readings for the Town of Cary using SAS analytics, enabling efficient data management, leak detection, and data-driven decision-making for conservation efforts.

Methodology

Integrated SAS for cleansing, visualization, and statistical modeling of over 500M readings from 60K wireless meters. Applied time-series and anomaly detection to identify usage spikes and missing values, with dashboards built in Power BI.

Key Finding

The SAS analytics system detected a major consumption spike on July 1, 2022, saving the municipality millions in potential water loss by automating leak alerts and optimizing billing accuracy.

Visual Analysis & Insights

Water Dashboard Overview
Dashboard Overview

Insight: Interactive SAS dashboard combining meter readings, water usage patterns, and leak detection summaries. Enabled decision-makers to view consumption trends in near-real time.

Water Data Overview
Data Overview

Insight: The dataset included six core attributes – TypeM, Time, Value, Unit, Series, and ManagedObjectID. Visualization highlighted distribution patterns and missing data intervals before cleansing.

Cleansed vs Uncleaned Data
Cleansed vs Uncleaned Data

Insight: Post-cleaning, the dataset achieved 0% missing values in "Time" and "Unit" attributes. This drastically improved the accuracy and stability of analytical models and dashboards.

Water Spike Analysis
Spike & Anomaly Detection

Insight: Analysis revealed a sharp anomaly on July 1, 2022, where total consumption surged beyond expected norms, signaling a likely leak event.

Statistical Summary
Statistical Summary

Insight: Data summary showed an average value of 13,990,876 liters across readings, with a minimum of -140 and a maximum exceeding 1.6 billion, confirming presence of outliers.

Time Distribution
Time Distribution

Insight: Time-series plots displayed improved continuity after imputation, reducing left-skewness and providing smoother hourly data curves across the 72-day period.

Unit Distribution
Unit Distribution

Insight: "L" (liters) dominated the dataset (621,968 entries), confirming it as the primary measurement unit, while "C" and "dB" represented sensor diagnostics.

Technical Implementation

Data Cleansing & Processing

SAS analytics tools were used to detect and replace missing or duplicate values. The unit column was imputed using mode replacement ("L"), while time values were filled using statistical interpolation for continuity.

Visualization & Reporting

Dashboards built in SAS and Power BI enabled dynamic visualization of consumption trends, leak alerts, and meter performance metrics across all zones in the Town of Cary.

Infrastructure & Integration

Data stored on cloud platforms (AWS/Oracle) for scalability and accessibility. Integration with municipal systems facilitated real-time alerts and customer portal updates.

Project Impact & Outcomes

Efficiency Gains

Saved over $10 million through early leak detection and automated meter readings.

Customer Empowerment

Residents can monitor usage and receive alerts, promoting conservation and awareness.

Sustainability

Supports eco-friendly initiatives through reduced waste and optimized water distribution.