— case study
Distributed energy resource (DER) systems offer several advantages over traditional centralized power stations. Aside from the potential long term cost-savings for asset owners, because such systems are typically installed where the power is needed, they reduce the inefficient loss of energy from transmission and distribution, while also supplying surplus power to the grid.
However, these systems are not without their limitations. Solar panel resource outputs, for example, are highly sensitive to environmental interferences such as dust, time of day/year, location and weather conditions. Failure patterns, and their appropriate corrective actions, are well known and documented but for utilities operators such as ZTRIC, a key challenge is the problem of effectively optimizing the energy production of a vast number of panels, and resolving issues promptly and cost-effectively.
As Pete Schleider, Chairman of ZTRIC explains, “Due to the large number of panels we monitor, our biggest difficulty was identifying when a problem had occurred. We needed a way of predicting potential system failures so we could take the necessary steps to prevent them — or better yet, we needed a way of predicting potential failures and taking preemptive steps to prevent them.”
“We needed a way of predicting potential system failures so we could take the necessary steps to prevent them.”
Pete Schleider, Chairman, ZTRIC
Through the use of analytics and Artificial Intelligence (AI), ThinkBridge built an operations and maintenance platform to allow utilities operators to gain greater visibility and control over their disparate solar resources.
Five components of the platform help ZTRIC better monitor the vast numbers of solar panel resources, anticipate environmental conditions, and accelerate the resolution of production interruptions and reduce losses:
A consumer self-service app that allows panel owners to view the performance of their solar energy system. It provides information on the health of the system, its production, cost savings and their contributions to a ‘green’ environment.
A cloud-based monitoring platform is used to gather live data from the sites and feed an AI engine.
Machine learning algorithms trained with both internal and external data help identify likely causes of production drops and create maintenance tickets. The system continuously learns ‘normal’ patterns and identifies anomalies that trigger alerts with a suggested plan of action. The algorithm also creates preventive maintenance action plans based on patterns of previous alerts to anticipate and mitigate the risk of future production interruptions.
An automated dispatching platform informed by the machine learning engine creates focused task lists and dispatches the appropriate technician. The dispatching platform then tracks the status of the tasks and feeds it back to the machine learning engine to verify if the issue was resolved.
An online marketplace of ZTRIC certified technicians and/or other support staff allows available technicians to pick up jobs from the dispatching platform, complete them and get paid.
By implementing ThinkBridge’s AI-enabled operations and maintenance platform to automatically monitor the health of all their systems, and automatically trigger notifications to dispatch the appropriate personnel, ZTRIC was able to dramatically scale its operations and provide a round-the-clock service and preventive maintenance guarantee to its customers (solar panel owners) ensuring the best performance and ROI of their solar panels.
In the future, ZTRIC intends to continue expanding its use of data and AI throughout its value chain to better integrate their resources, and make quicker, more informed decisions to optimize the performance, reliability, resilience and security of its business and energy grid.