Building a Knowledge Infrastructure for Utilities

Preserving institutional knowledge

When I ran an electric utility operations division, one of my favorite employees was a guy named Stanley. Stanley started as a line worker; climbed poles; became a foreman, later a supervisor; then managed all the crews in the region. I remember how Stanley worked.

As the hot, humid day would turn into evening, just when the crews returned to the service center, storm cells would start to form. If they matured, they could cause heavy rain, wind, thunder, lightning, and sometimes minitornadoes. Stanley had to decide whether to send the crews home or keep some or all the crews on overtime. No one really knew if the storm cells were going to dissipate or cause havoc to the electric system. If Stanley sent the crews home and a bad storm hit, it would take a long time to get the crews back to work. If he kept the crews on overtime and the cells dissipated, he would have wasted company money. Stanley almost always made the right call. He didn’t know it, but he was using spatial analytics in his head.

Then Stanley retired.

The average age of U.S. utility workers is almost 50. Thousands of workers like Stanley will leave the industry over the next several years. Imagine all the wisdom and analytic power that will be missing. People like Stanley know where infrastructure problems exist. They know where the utility has not trimmed trees. They know the location of old and frayed wires that are just waiting to fall down. They remember where storms generally hit and the problems storms cause.

What many utilities are missing is an ability to capture as much of that wisdom as possible before the Stanleys of the industry retire. What we need is a way to share what retiring workers know and how they know it. The common denominator of that knowledge is location. Utilities have been capturing facts in geographic information systems (GIS) for years. Today, GIS can capture observations and predictive information, collect data from all kinds of sources, and help utility staff make better risk predictions the way Stanley did. GIS can create geoprocessing models, which document the data sources, run the analysis, and produce the results in the form of a map. The key is to have these models validated and supplemented by experienced workers before they leave, so that utilities can truly build a knowledge infrastructure.

Can the utility GIS community provide a platform to build a knowledge infrastructure that leverages experienced workers before they leave?

Bill Meehan

About Bill Meehan

Bill Meehan, P.E., heads the worldwide utility industry solutions practice for Esri. Author of Enhancing Electric Utility Performance with GIS, Modeling Electric Distribution Performance with GIS, Empowering Electric and Gas Utilities, Power System Analysis by Digital Computer, and numerous papers and articles, Bill has lectured extensively and taught courses at Northeastern University and the University of Massachusetts. Bill is a registered professional engineer. Follow Bill on Twitter.
This entry was posted in Industry Focus and tagged , . Bookmark the permalink.

Leave a Reply


  1. Although this post is directed specifically toward the GIS utility community, this is an issue along all aspects of government. As GIS manager of a large metropolitan county, this issue is seen across almost all departments. It is even more pronounced in those agencies who have only just begun to embrace GIS technology. How do we capture the years and years of experience in emergency operations, traffic management, 911, elections, etc. before these experienced workers retire?

  2. Thought provoking post, Bill. Thank you. As rightly pointed out by Victoria Morrow, this is an issue across all departments that collects, and stores data. This institutional data is a must in almost every industry. Some industries overcome this challenge to some extent from the knowledge acquired through historical incident data.
    Institutional knowledge management is similar to meta data management, a challenging workflow in several federal, private agencies, usually blamed as a time/resource consuming process. How do we capture this institutional knowledge? Well, combining old school methods and latest technologies could help. Imparting institutional knowledge collection in the data collection workflow is a way to go. Better late than never! Also, exploring innovative ways of emerging augmented reality technologies + GIS + temporal data is another possible way to capture those years of experience from Stanleys.

  3. Ron Brush says:

    This question seems to come up more frequently and I think it’s important. Capturing Subject Matter Expert (SME) data is a challenge but there are solutions.
    One approach that we have used successfully is to gather some of the SME’s in a room. With a meeting facilitator and GIS Analyst and with a system map on a big screen, we ask the SME’s to tell us about different parts of the system they know best. The analyst creates new features and feature classes – usually polygons, and captures their knowledge about that part of the system. This can include known problem areas, installation methods and materials that were used, landowner information and much more. The SME’s name, date and other metadata is also attached. This information can then be vetted with other SME’s and later be organized into a more usable format.
    This spatial SME approach will be important for gas distribution operators as they move forward with their DIMP planning and implementation.
    While this approach may not replace Stanley’s ability to predict the weather, it can help retain valuable knowledge about utility assets that might otherwise be lost. Plus I think it’s a complement to the SME’s to acknowledge their experience and value to the organization.

  4. Bill Meehan Bill Meehan says:

    Great comments folks. The idea of gathering experts in a room and drawing their concerns on a map for entry into a GIS has lots of value. Using geoprocessing (spatial analytics) this pseudo-authoritative data can be combined with authoritative data to connect the dots, between what people know in their heads or remember from experience with stored data from say a maintenance management system or with data from historical prediction data (like lightning strike data). Also the visual presentation of model builder gives experienced workers the ability to vet whether all the variables have been considered say in a risk assessment. Stanley might look at the model and say, you forgot to include soil types in the assessment, since trees fall over easier in sandy soil than in clay.

  5. It is encouraging to hear that companies are seeking ways to capture our Subject Matter Expert’s (SME) pools of experience and knowledge in a GIS environment. We have spoken of this concept in our shop as well and I think it could be worthwhile to pursue. As a GIS Technician with skills in building Geodatabases, spatial analysis and modeling; I would be pleased to add this type of endeavor to my project list! In this discussion, it would be great to hear some more specifics in terms of data formats, metrics, features and attributes that seem of value and could be collected, compiled and archived during sessions with SME’s on areas of interest/concern to electric utilities. What questions do we often need answered by people with long term experience in an industry?