Norway-based TOMRA is driving a more sustainable future with its sensor-based systems to scan and sort products and materials for the food, recycling, and mining industries. To increase workflow efficiency and perform quicker queries on massive data volumes, it transitioned from a monolithic SQL server infrastructure to Azure Data Explorer (ADX). With the new solution, TOMRA can process up to five billion data points per day, enabling more machines to identify and sort materials for recycling quickly and accurately.
Leading the way in sensor-based sorting and processing technology, TOMRA services the food, recycling, and mining industries around the world. Using Internet of things (IoT) sensor technology, the company collects and analyzes data from hundreds of food sorters and recycling systems to sort different materials efficiently.
"Each of our machines generate about one million data points daily that need to be processed quickly. However, our legacy SQL server that compiled the sensor data was only connected to five percent of our sorting machines’ data outputs,” says Christian Wuerdig, Lead Cloud Architect at TOMRA.
Additionally, the query performance resulted in frequent timeouts. The SQL environment was also difficult to scale quickly to accommodate the growing number of machines and sensors. “In the next few years, we want to have at least 5,000 machines connected. So, we needed a platform that could handle five billion data points a day in the future," says Wuerdig.
The cumbersome, monolithic SQL server needed to be replaced with a solution that can query different data formats with speed and ease. Wuerdig explains, "It wasn’t easy to set up the SQL server to manage more data points, so we started looking around for more adaptable technology.”
Huge data volumes at double speed
TOMRA decided to implement Azure Data Explorer (ADX) analytics. Working with technology partner Inimco, TOMRA swiftly deployed the new solution that has since streamlined the company’s workflow.
"The biggest win for us is the saved time. It’s so much easier and faster now to connect the server to some legacy or new unit. The new setup is also much simpler—you just turn it on and dial-up," says Wuerdig. The cloud-based system can also dynamically scale when required and is easier to maintain than on-premises systems. "We can now focus on other important aspects of the business," he explains.
The ADX solution has also helped save money for TOMRA. "We've reduced about 80 percent of our maintenance overhead costs," says Wuerdig. "The new infrastructure will probably save us about 50 percent in running costs and operational expenses since we can now run twice as many queries," he says.
TOMRA has paved the way for further innovations, such as using artificial intelligence to further enhance foreign material detection in food production lines. The company is also building Azure-based replicas of their customers' factory facilities to reduce downtimes and incidents. Getting firsthand data and feedback keeps TOMRA ahead of the curve. “I think it’s important to try new solutions and get customers’ feedback. It helps us understand what our customers need, which directs our plans for future technology,” concludes Wuerdig.
“The biggest win for us is the saved time. It’s so much easier and faster now to connect the server to some legacy or new unit. The new setup is also much simpler—you just turn it on and dial-up.”
Christian Wuerdig, Lead Cloud Architect, TOMRA
Follow Microsoft