UtilityServicesCo automates correction of missing and faulty smart water meter readings with VEE system
UtilityServicesCo, based in St. Louis, Missouri provides water, gas and electric Utilities with Smart Infrastructure Solutions. They empower 1000+ global Utilities with visibility over their distribution networks which advances their function and value. Their comprehensive suite of solutions consists of meters and edge devices, advanced metering infrastructure (AMI), consumer engagement software and provision of labor for installations. Their data-driven solutions help Utilities to predict, plan and respond to system conditions. In 40 years of existence, they were trusted for reliable and modern solutions. But UtilityServicesCo’s legacy systems reduced their likelihood of achieving the ultimate goal- to conserve more and more energy every day. Here are the limitations they faced-
No system to automate data validation, estimation and edit
UtilityServiceCo faced challenges with automating the validation, estimating and editing of the water data collected from 1.2 million smart meters worldwide. For their operational staff, validating such a monstrous amount of data in Excel sheets, was impossible. This process was especially a crucial one as it directly affected the efficiency of water bill generation, water consumption patterns trends and predictions related to meter conditions. Without an automated system, there could have been gaps in the smart water meter readings, no way to correct them, and generate accurate water bills out of them.
No system for bill corrections
UtilityServiceCo’s Utility customers faced the likelihood of generating false water bills for their consumers. This was due to the humungous amount of usage data collected by smart water meters at intervals of 15 minutes, throughout the day. In case of meter failure, the readings used to be mistaken; such as usage spike (excess consumption reading), zero reading (no consumption), static (consistent reading for several intervals), negative (less than zero consumption). The similar case of mistaken readings was applicable to readings collected over a period of days together. Generally, consumers showed similar patterns in water consumption over weekdays, weekends, holidays etc. This resulted in false calculations in water bills. This could have reduced the customer confidence in Utilities and also loss in revenue, if/when they refused to pay for false amounts.
Increased turnaround time for customer queries
The end consumers i.e. residential, commercial and industrial water users would complain of their inaccurate water bill to upstream Utilities. These Utilities then had to check their records and correct the bills in such cases. They had to physically check the meters and their data. After checking the readings, they had to compare the present & historical data and estimate the correct readings. This was a time-consuming task. It would often result in a high lead time to address and resolve customer queries.
Increase in outstanding bills
In case of generating inaccurate water consumption data, Utilities faced chances of increase in outstanding bills and delayed payment received from their customers. Usually, the cause of the inopportune bill amount would reflect itself in the meter data. Its careful analysis, comparison with historical data and rectification would happen manually. It was time-consuming and created lags in receiving payments.
Challenges with predictive analytics for smart water meters
The Utilities could not execute preventive and predictive analytics for their smart water meters. This was due to the absence of automation for data analysis. The discrepancies in the meter data were not sorted, identified and grouped together. It made data correction cumbersome. This inaccurate data created roadblocks with the future maintenance of the devices. The water meters’ productivity, operational efficiency of Utilities and customer confidence for UtilityServicesCo was at risk here.
Increased water loss
UtilityServiceCo was unable to fulfilling its goal to conserve water due to the absence of an automated data management system. With no actionable insight over its water distribution network, they faced issues with knowing meters’ health. They could not regulate future water supply in localities and bridge the demand-supply gap. With no knowledge of meter failures, leakages, outages etc. they were prone to water loss. There was a strong likelihood of consistent water loss in case of static readings (consistent readings for several intervals at once).
Solution: Automated Azure based VEE (Validation, Estimation and Edit) System for water meter data management with data analytics
Saviant built an automated Azure based VEE (Validation, Estimation and Edit) System for UtilityServiceCo’s water data management. This system gathered data from smart water meters. UtilityServicesCo collected, observed, stored and managed this data on cloud and eliminated Excel. It helped them automate data processing. Today, they can process 1 Bn+ data records per day coming in from 1.2 Million smart meter devices. The VEE System used trained algorithms to process the collected data. It was then validated i.e. reasonalized by comparing it with historical data of the same account. Other methods based on other rules which did not include historical data were also created. Consumption patterns were analysed and holidays, weekends, weekdays etc. were taken into consideration to create rules. These rules helped identify discrepancies in data and estimate the corrections. These corrections were then implemented, and accurate data usage was calculated. IT helped Utilities generate accurate bills for its customers. Their customer confidence increased. Due to the real-time visibility over its smart water meters, the peak usage periods were observed which helped with water conservation, motivating consumers towards water conservation.
Solution Architecture Diagram
Need authentic meter data for improved operations?
Talk to our Analytics Expert