Data DiversityAugust 12th, 2016 by Matthew Ross
The Energyworx Approach
In our last few blogs, Energyworx discussed Cloud Machine Learning and viable use cases for the Energy Industry. We now take a look at where the data is coming from and how to ingest, crunch, learn, and monetize it.
The rollout of smart meters has triggered the term smart grid but we know the smart grid incorporates far more resources than meters alone. It’s a good thing too as the promises of the smart meter would largely be unfilled without additional resources to enhance their value.
To strengthen our customers’ use cases, Energyworx ingests time-series data from many other smart grid resources such as smart inverters, EV charging infrastructure, IoT devices, smart thermostats and grid sensors (uPMUs). In addition, Energyworx collects relevant metadata to complement the time-series: customer characteristics, social/demographic info, contract details, market/pricing info and equipment specifications. As a result, Energyworx’ version of the smart grid is much more diverse and data rich, requiring advanced energy data management and intelligence solutions to truly monetize the smart grid.
How do we interpret all these datasources?
Much attention is given to the sheer volume of data the smart grid is generating and for good reason. In a previous blog post, we calculated that the growth in meter data alone is over 30,000x. From Energyworx’ beginning, we’ve been working with more efficient and scalable architectures for ingesting, storing and processing time-series data. In addition, we’re utilizing Pure Play Cloud Computing so we can scale up or down based on the given demand.
Now turning to data diversity, traditional software vendors struggle to keep pace with all the innovation going on in the smart grid. Many of these vendors are supporting multiple versions of their software for their different customers. This means that a new (IoT) product or protocol that enters the market must be incorporated into all versions without creating any “breaks” in the software. This is a slow, tedious process and may not even be offered by the vendor until the next software version is available.
Energyworx instead works with a single baseline of code, meaning there is only one version of our software deployed in a multi-tenant environment. The software is updated every two weeks with new features that all customers can instantly utilize. This strategy brings incredible economies of scale and ensures Energyworx and its customers can keep pace with market changes and thrive against the competition.
Energyworx ingests data in many different ways but always with the same approach. Energyworx implements data conversion adapters that are capable of converting a specific protocol to a protocol required for Energyworx. Energyworx has a variety of adapters for different formats and creating new adapters typically requires little effort (days or weeks). Many adapters are available “on-the-shelf” including the leading smart meters, headend systems, smart inverters and IoT products. Once these adapter are developed, Energyworx can ingest data and begin its validation, editing and estimation (VEE) engine. This approach allows our customers to ingest raw data, check the quality and improve it if necessary.
Your applications are only as good as the data you are using. Make sure you have access to all smart grid datasources with the highest data quality possible.
How is your organization dealing with data diversity? Contact us to see if the Energyworx Approach can improve your business case.