Optimizing Maintenance of Digital Billboards



A large digital display company designs, manufactures, sells, and services digital billboards that you often see on the roadside.  Like all electronics, the company’s products are prone to failure and require maintenance from time to time, but they struggled to service their displays in an efficient manner, leading to high warranty and service costs. Diagnostic data, collected from their displays, was leveraged to predict future failures, and a business strategy was developed to proactively service products.




Hundreds of gigabytes of diagnostic data from thousands of active displays was pumped to an Azure Data Lake every day, taking advantage of the Internet of Things. Using Azure Data Lake Analytics, Azure Data Factory, and Azure SQL DB, consumable data was made accessible to users in the business.  In addition, Power BI reports and dashboards were developed to meet regular monitoring and analysis needs, while Azure ML was used to leverage machine learning algorithms and deploy predictive models that informed the business of probable product failures in the future.


Key Outcomes


We collected hundreds of gigabytes of data to inform the business of probable product failures so that the displays could be maintained more efficiently. While this project is still being deployed, the company expects to identify opportunities to improve future products, proactively service current products, and dispatch service teams more efficiently. Over time, these improvements will lead to a decrease in failures, increasing customer satisfaction, and allowing the company to retain more customers while simultaneously reducing warranty and service costs.