Client background

This client manufactures high-end kitchen appliances such as refrigerators, freezers, ranges, cooktops and grills in their American production facilities.

The business challenge

The client had recently implemented connected sensors in their production equipment, allowing them to capture real-time data on temperature, humidity and other readings at various points in the manufacturing process. The implementation of two central data stores for telemetry allowed them to collect massive amounts of data. However, analysts and engineers couldn’t use the data effectively in its native format or surface the information in real time using existing canned reports, requiring complex transformations to extract insights. The presence of multiple databases also meant that any analytics solutions built directly on top of the databases would have to be duplicated. Additionally, because the client’s existing legacy On-Line Analytical Processing (OLAP) cubes were no longer able to support the continued growth in size of their data warehouse, they were planning a long-term shift to the cloud.

Strategy and solution

Baker Tilly Digital worked with the client to create a solution critical to the client’s cloud architecture that included:

• Real-time sensor readings at a detailed and aggregate level with low latency
• Azure Event Hubs to push data from the on-premises telemetry databases into the cloud
• Processing and transforming data into a single real-time feed using Azure Stream Analytics
• Leveraging the integration of Power BI and Stream Analytics to display relevant metrics in live visualizations
• Tabular models to produce more responsive reporting than legacy OLAP cubes for both executive and operational reports
• Automation of the conversion process to significantly shorten the timeline of the migration, freeing up the client’s team to work on converting business-critical reporting
• Automated daily maintenance to keep costs down by eliminating spend outside of business hours
• Incremental refreshes to reduce data transfer, increasing the team’s ability to meet service-level agreement (SLAs) and reducing cloud costs

Baker Tilly’s approach allowed the client to achieve:

• Real-time view into the health of their production process, enabling them to flag equipment approaching critical limits before problems occurred
• Observation of long-term trends in machine health at levels of detail previously unavailable
• The ability to make data-driven decisions in the improvement of their production process while allowing the client’s internal analytics team to focus on delivering value directly to the business
• Time savings, optimized cloud costs and adherence to their larger cloud migration plan