Bandera Electric Cooperative Develops Smart Energy Analytics Solution Using AWS IoT

Millions of people pay for their electricity each month without knowing how their energy usage affects their bills. To remedy this, Bandera Electric Cooperative (BEC) developed Apolloware, an energy analytics solution that helps users monitor their home energy usage data. However, BEC’s original on-premises architecture was inefficient at streaming data, resulting in frequent system outages. By migrating to a serverless Internet of Things (IoT) architecture using Amazon Web Services (AWS), BEC improved the stability and security of Apolloware so that it could offer the solution to more consumers. Apolloware users can now monitor their home or business energy usage through a mobile app or website and use this visibility to reduce their energy consumption to save money.

Established in 1938, BEC serves over 27,000 members in central Texas. It launched its solar division in 2017. The cooperative is focused on providing quality service for its members while promoting sustainability and grid resiliency.

As part of its mission to empower consumers to make smarter energy choices, BEC identified a problem. Its members had no way of knowing how they were consuming the energy being delivered to their homes. To address this problem, the cooperative developed Apolloware. Unlike smart meters, which simply reduce the need for a meter reader, Apolloware lets consumers monitor their energy usage to take control of their power consumption, resulting in cost savings for them and lower carbon emissions for the planet.

BEC developed Apolloware in 2016 using a single-board computer connected to consumers’ smart meters. This early solution streamed data to databases on BEC’s on-premises servers. However, the servers weren’t efficient at streaming data, and BEC’s databases needed constant oversight. “Our original on-premises SQL databases were architected for a human to observe and extrapolate data from,” says Justin McKenzie, technical lead for Apolloware at BEC. As the cooperative sought to expand its use of Apolloware, it found that its on-premises system could only handle data from 58 sites before experiencing system outages. It was then that BEC began to look for alternative ways of developing Apolloware to meet its plans for current and future growth.

BEC needed an affordable, scalable solution—one that would let it organize and restructure the flow of data from its smart meters. It found this solution in 2019 when it began using Amazon Elastic Compute Cloud (Amazon EC2), a web service that provides secure, resizable compute capacity in the cloud, and Amazon Redshift, a cloud data warehouse that makes it easy to gain new insights from data. BEC can now process near-real-time telemetry data across all utilities and can meet utility consumers’ demand without limiting the number of members. “By going from on premises to a scalable cluster solution, we can evolve to handle scale,” says McKenzie. “Using AWS, we can manage our hardware needs and, ultimately, the foundational database that supports Apolloware.”

To collect energy consumption data from its smart meters, BEC uses Amazon Kinesis Data Streams (Amazon KDS)—a massively scalable and durable near-real-time data streaming service—to ingest data from its meters to the cloud. Apolloware’s mobile app shows users the total consumption of energy in their homes and then breaks that down to show how much energy each appliance is using in near real time. To process and transform the ingested data, BEC uses Amazon EMR, which makes it simple to set up, operate, and scale big data environments by automating time-consuming tasks like provisioning capacity and tuning clusters. Apolloware also notifies users of their projected energy usage for the day, along with suggestions for how to plan and budget that usage, by using AWS IoT Events, a fully managed service that makes it easy to detect and respond to events from IoT sensors and applications. The app provides alerts for events such as heat advisories and rolling service outages and has proven vital to communicating with users during winter storms. “AWS IoT Events is a critical piece of our alert system. We can use it to educate our members on how they can help support the power grid during storms,” says McKenzie.

Using AWS IoT Events, BEC can detect when equipment on the grid is about to fail—information that it uses to prevent or minimize service outages. The cooperative is also using data gathered from its smart meters to develop predictive analytics for grid maintenance. By managing its grid infrastructure with continually improving data, BEC improves system reliability while also helping users reduce their energy consumption by an average of 12 percent. Due to the insights provided by Apolloware, over 150 members have made targeted improvements to their homes—such as weatherization and heating, ventilating, and air-conditioning upgrades—that have resulted in cost savings of over 20 percent on their monthly energy bills. One member saved 20 percent on his bill and reduced his energy usage by 26 percent compared to the same month of the previous year. During the same time period, another saved 35 percent on his bill and reduced his usage by 41 percent.

In 2020, BEC installed Apolloware on all the solar systems on its grid, and the cooperative has plans to offer Apolloware to other utility and solar companies starting in 2022. Using Apolloware, utility companies gain a better understanding of their residential solar resources and the impact to grid equipment during peak times. They also gain greater visibility to meter appliance usage at the fleet level—for example, seeing how electrical vehicle chargers are impacting the grid in near real time.

To improve alerts and engagement, BEC uses AWS Data Lab, which offers accelerated, joint engineering engagements between customers and AWS technical resources to create tangible deliverables that accelerate data and analytics modernization initiatives. As Apolloware continues to scale, it is evaluating solutions that can drive efficiency, performance, and cost savings. The cooperative looks forward to using Amazon Timestream—a fast, scalable, and serverless time series database service for IoT and operational applications—to further reduce costs and enhance the performance of Apolloware as it continues to implement AWS machine learning solutions. “We’re making jumps in technology to keep us on the cutting edge,” says McKenzie. “The next phase of this is to determine trends and characteristics of the data.”

BEC plans to use the machine learning framework of AWS Data Lab to provide recommendations for customers, such as modifying automated settings on a pool pump or changing an air conditioner filter, along with estimated cost savings for these changes. “Our belief is that when you start looking at those little bits of data and make them visible, homeowners are going to make the right choice for their families, their budgets, and ultimately the reduction of energy in our territory and other territories across the United States,” says McKenzie.