The % use of predictive and prescriptive analytics in the manufacturing plant can greatly increase capacity utilization, reduce plant downtime and optimize costs. A leading manufacturer of high-performance materials and specialty chemicals needed a solution to predict optimal machine performance, ensuring that unscheduled downtime didn’t cause losses. Their current legacy IT infrastructure was unable to predict machine breakdowns for more effective preventive maintenance.
The client needed advice on what technology to use and a partner to evaluate different AI platforms and cloud providers to meet their business requirements.
Preparing for fault-free production: Collabera reviewed their existing infrastructure and recommended Microsoft Azure’s Data services as the foundational AI/ML platform. Our solution included defining the data architecture and setting up the data and analytical services for real-time sensor data ingestion, processing, anomaly detection, regression prediction, and visualization.
As a first pass, the solution was set up to bring data from the filters that were used in the client’s extrusion process. These filters were high-value sub-assemblies that required accurate failure prediction so that optimized inventory was maintained to ensure minimal downtime to this critical production process.
- Designed an IOT solution to capture plant and machine data from the factories, transform the data, and store it in Azure Data Lake (ADL)
- Configured the solution and established connectivity with data sources in 8 weeks
- Ensured proper ingestion of accurate data into the PowerBI reports and Azure Auto ML for “CP Filter lines” to predict faults
Once the concept was proven, we worked on a plan for loading data from 80,000 sensors from one factory into ADL. The solution is aimed at transforming the data so it can be consumed by the data science team for analytics solutions across various machines covering over 25 use cases that were jointly identified.
Azure Data Lake, Azure VM, Azure IoT hub, Aspen, SAP, SQL managed DB
Collabera developed the AI/ML platform infrastructure and constructed an implementation roadmap for the client. The data architecture and management strategies used resulted in an increased lifespan of the equipment by 6 to 8 months and many other successful results:
savings due to the increased machine lifespan
savings per month due to improved utilization of raw materials
loss prevention from production unit downtime
predicted savings for each of the 25 planned business use cases
The production helped to predict failure and process issues while also providing proactive maintenance leading to higher quality, faster time-to-market, and optimized costs.
Collabera has co-engineered and helped numerous organizations succeed in their architecture modernization efforts. Talk to us to learn more.
Ready for the next?Talk to experts
Get the latest updates
Request for Services
Collabera developed the AI/ML platform infrastructure and constructed an implementation roadmap for the client. The data architecture and management strategies used resulted in an increased lifespan of the equipment by 6 to 8 months and many other successful results.