When the power grid fails, the failure can be life threatening. The Resilient Reliable Efficient Energy Systems (RREES) Lab aims to deliver power quality data fast enough to keep the failure from becoming an emergency, detailed enough to identify the failure, and comprehensive enough to locate the failure.  While the Smart Energy Analytic Disaggregation System (SEADS)  handles the collection of an overwhelming quantity of data ( ~5.5GB per day per device uncompressed), our team built the SEADS Vault to:

  • identify important events in the SEADS data, 
  • notify homeowners of events, 
  • inform grid managers of event locations 


How Outages are traditionally handled:

  • A User calls their provider to manually report.
  • Perform a scheduled/monthly maintenance.
  • Voltage is measured at the substation

How we want to change it:

  • Automated Event Detection and notification
  • Isolate problem areas and fix faulty equipment
  • Voltage is monitored at the consumer’s house

All these factors provide better system health:

  • Monitors on the Consumer level leads to higher accuracy and can be used to link common events to or faulty equipment.



Architecture Diagram

Architecture Diagram

Technology used includes React, Flask, and PostgresQL for development. We incorporated as third party services, such as Auth0, Mapbox, and Twilio; in order to handle account management, geolocation, and notifications. Finally, we used Skyhook for partitioning on each SEADS Device.


Downed powerlines

Downed power lines can cause safety issues like wildfires.
Fallen tree

Here, a tree falls down in an urban area and is not detected for a long time. This causes a power outage and a road obstruction.

SEADS Vault Features

1. Have an application deployed to the web.
2. Display real time live voltage to end user.
3. Notify consumers of potential outages in interested areas via email and text.
4. Flag defective equipment for repairs ASAP to improve safety and improve reporting.


We hope to establish a strong foundation for future innovators to build new features using SEADS and extend the user base outside of Santa Cruz. For instance, researchers and scientists could use  the data to develop machine learning models.



Brian Zhao: Phase 1 Design
Patrick Mantey: Mentor and Sponsor
Jeff LeFevre: Creator of Skyhook and Design
Michael Choi: Building the SEADS Hardware
Faeze Brahman: Managing Weekly Scrum
Akila De Silva: Managing Weekly Scrum
Richard Jullig: Bi-Weekly Check instructor