The whole team has been working hard on TGI over the past few months – both on the front end and under the hood.  We’ve been listening to customer feedback as well as the experiences of our own technical services team to prioritize development efforts.  As always, our goal is to make TGI more effective and TGI users more productive.  Here are the main highlights of 2.6:

Enhancements to Risk Modeling

We’ve added a new interface to help data scientists/analysts select the data sources they want to use for building risk models.  The new screens allow users to choose data source containing the risk factors, apply weights to these factors and combine them to create their risk models.

We’ve also added new tools to analyze the effectiveness of risk models using the real-world data collected during investigations. Users can visualize how the risk ranking that was produced by their model compared to actual amounts of loss found. By feeding back actual measurements into the model, TGI gets progressively better at finding losses.

Improved Field Investigation Tools

We’ve added a great new feature for investigators that traces the connections between grid elements and shows them on a map.  European crews will be especially happy as they can now see these connections in a mesh grid environment where there might be multiple feeders in one circuit or loops.  We’ve also added the ability for crews to apply corrections to switch states on their TGI map when they find things not agreeing out in the field.

New Program Dashboard

Our new dashboard provides an overview of investigations planned, in-progress and completed as well as the lineage of the investigations.  Key KPIs like investigations by type, percentage of grid investigated, success rates of investigations, technical vs non-technical losses, and other indicators are available:



If you have any feedback on these new features, or if you have any suggestions for upcoming release, please let us know!