August 19 Grid Analytics Community – Machine Learning in the Field: Optimizing Meter Gateway Locations

Join the Grid Analytics Community on Wednesday, August 19 at 1:00 PM CT for a presentation with Q&A/open discussion on “Machine Learning in the Field: Optimizing Meter Gateway Locations” led by Abigail Heinsohn, Data Scientist at Delaware Electric Cooperative.

Session Description:

For the presentation “Machine Learning in the Field: Optimizing Meter Gateway Locations”, DEC’s engineering department wanted to add 5 new gateways to the system but wanted to focus on areas that had meters with high hop counts. Higher hop counts indicate the data has more channels to jump through to get the information back to Yukon which can cause the data to arrive slower or possibly not at all. In addition, each gateway can only accommodate 5,000 meters. The data department was tasked with using the data to find the optimal locations to add the new gateways. Using a ball tree nearest neighbor algorithm in Python, they were able determine these locations based on longitude, latitude and hop count of current meters. After adding these gateways to the system, they looked at the before/after impacts to the hop count and identified the meters/locations impacted the most. They were also interested in areas that still had high hop counts to recommend during the next iteration of gateway installation. Lastly, they checked for redundancy to see if meters are dependent upon one gateway or have nearby options as there are sometimes issues with gateways going offline.

*UAI Utility Members can register and access all nine of our Analytics Communities via our Community Page. Not a part of the Grid Analytics Community and want to join this session? Please use our direct link here to request access. You will find the Teams meeting credentials to the upcoming meeting at the top of the feed on this page. If you have any questions or run into any issues, please contact us at info@utilityanalytics.com.