Case Studies

Utility Analytics

A Canadian Utility company wanted to build an analytic model to predict the health of their assets in order to reduce the cost of manual inspections. As part of this model building exercise Alteris analytics team was also tasked with classifying and re-coding within the Utilities ERP systems directional comments written by inspectors to add depth to the modelling.

 

Solution

  • Formulated a set of  logical/non logical criteria sets as a frame work to define and determine the outcomes with the data.
  • Itemized criteria that would take into consideration location and building demographics to determine a series of metrics for health and condition.
  • Upon choosing the “first” set of criteria blocks and site location, a multi Geo-spatial view of the location and itemize details we’re captured in the same area within a specified agreed to radius with the Utility to focus on quick wins.
  • A sequenced approach was taken to deliver outcome results on age data asset from the oldest and delineate asset detail with a clear output on the “subdivision” as to be defined.
  • Concluding a  bench mark evaluation of the existing infrastructure against reconstructed areas that have taken place and outweigh that baseline to our defined criteria as a baseline calculation on outcomes.
  • Comprehensive and propriety analytic utility modeling tools we’re leveraged by Alteris, to provide a robust predictive modeling solution that effectively delivered cost management activities, operational efficiency actions  and safety metric impacts for effective Utility Asset management.

Value Achieved

Based on the categorization of the inspectors’ written comments, Alteris was able to include valuable information in health asset model building for effective asset planning. This user generated data was one of the most significant variables having great impact on model predictability.

 

The key accomplishments include:

  • Providing the utility provider for the first time the ability to look at this data stream in an aggregate format
  • Allowed for better assessment of type of damage to their assets, allowing for better management of issues
  • Helped in refinement of next year’s budget; 20% of all issues were related to an itemize list of external factors
  • Improved ability to cross reference issues by geographic locations for better identification of areas prone to specific problems i.e. vandalism
  • Continued categorization of such data (over time) would allowed for identification of trends, seasonality linked issues and other patterns, which allowed for better preventive maintenance, reducing cost due to breakdowns