Landis+Gyr has partnered with US-based Grid4C to provide advanced AI and machine learning solutions for the energy sector. A joint pilot project in Australia has already shown promising results. In our interview, Shane Fay, Senior Vice President Global Sales at Grid4C, talks about how utilities can utilize these solutions to leverage their AMI data in a variety of business areas.

pathway: How can AI and machine learning help
improve business processes of utilities?

Shane Fay: In our partnership, Landis+Gyr and Grid4C leverage AI and machine learning to enable energy providers to maximize business value out of smart meter data and IoT devices. We deliver accurate, granular operational and customer-related predictions to improve operational planning and load forecasting, reduce peak demand, increase energy savings and optimize demand response (DR). We also help tackle challenges like distributed energy resources (DER) integration, customer retention and customer engagement.

pathway: What are the key areas of Grid4C’s predictive operational and customer analytics – and which insights do they provide?

Shane Fay: Grid4C’s solutions focus on three areas: customer-facing applications that help businesses and consumers not only to save money but also predict problems with the appliances they rely on, predictive customer analytics that facilitate segmentation and micro-targeting, and predictive operational analytics that predict grid asset failure, and optimize the integration and coordination of solar, energy storage and electric
vehicles.

The AI engine uses smart meter and IoT data to model each meter and endpoint and predict its individual behavior. The models disaggregate and predict usage for appliances behind the meter and are aggregated to deliver predictions for grid assets. By building predictions from the most granular level up, the core technology drives applications ranging from forecasting and optimization of distributed energy resources, to predicting, detecting, and diagnosing faults and inefficiencies for grid assets and home appliances, without the need for hardware or sensor installations. pathway: How can this information be monetized?

Shane Fay: The insights that our AI engine are able to extract from smart meter reads enable utilities to target offerings very precisely. In compliance with data protection legislation, we identify customers likely to churn and customers likely to adopt solar, purchase electric vehicles, or participate in other energy efficiency or marketing offerings.

Another example of how utilities can generate new revenue streams is by monetizing the unique capability to predict, detect and diagnose faulty appliances within the home using AMI reads only. This enables utilities to provide new value added service offerings for residential consumers, which alert customers when they have faulty or inefficient home appliances, quantify the costs of the appliance inefficiency, and guide them to buy relevant products and services.

pathway: How is the solution implemented into a utility’s process and system environment?

Shane Fay: We offer plug and play solutions, often needing less than two months to deploy them at scale. Customers can either use our displays and portals for a utility-branded experience, or opt for Data Science as a Service (DSaaS). In that case, we receive smart meter data and deliver the models and insights back to the utility to be displayed on their own screens or products.

pathway: Briefly: who should take advanced AI and machine learning solutions into consideration?

Shane Fay: Utilities and retailers who are interested in increasing revenues, maximizing profitability, enhancing customer satisfaction, offering value add services to their customers, or extending the value of their AMI investments are perfect candidates for our offerings.

pathway: Please share what the future looks like and how Landis+Gyr and Grid4C are aligned to deliver value to utility clients.

Shane Fay: The key trends of the industry are decentralization, electrification of private transport, increasing competition, growing intelligence at the edge of the grid as well as the integration of multiple products and services in connected homes. All of these will enable and require granular, accurate forecasts of energy consumption and production based on metering data. Landis+Gyr and Grid4C have the right solutions to maximize value for utility clients in the future.

in a nutshell: IoT predictive analytics
Landis+Gyr and Grid4C offer a complete suite of advanced tools for predictive operational, customer and home analytics. Machine learning and metering data are used to generate revenue streams, meet energy efficiency goals and to strengthen the market position in the highly competitive retail market. The solution allows accurate, granular predictions based on normalized data sets, APIs and signed meter data and is now close to mass deployment. “The simplicity of being able to go mass market with only the meter data is a very attractive offering for retailers,” as Rodney Chaplin, General Manager Australia & New Zealand at Landis+Gyr, puts it.

In the course of the joint project, a major utility in Australia reports:

• a reduction of the customer churn rate from 10% (customers not receiving insights
how to optimize their energy consumptions and the appliances they rely on) to 2.9%
(customers receiving these insights),
• a ratio of 57% of customers participating in the trial who have filled in their home
profile (in most cases with complete information about their home e.g., data that is
suitable for personalized marketing).