Fix the problem before it is a problem: predictive maintenance
- Predictive Algorithm
On average, a cow produces 6-7 gallons of milk a day. That milk needs to refrigerate, pasteurised, and stored in temperature-controlled environments. A breakdown in the operations can result in spoilage. The Global Academy of Agriculture and Food Security found that one in six pints of milk produced was wasted. This amounts to an enormous expense for clients of Tetra Pak Services, a world leader in process and packaging solutions for the F&B segment.
Tetra Pak is the world's leading food processing and packaging solutions company working closely with our customers and suppliers to provide safe food.
Tetra Pak Services launched a Condition Monitoring service to help food and beverage manufacturers predict machine failures before they occur.
Enter predictive maintenance. Working with Microsoft Azure, Tetra Pak Services introduced the Industrial Internet of Things (IIoT) to their machine line. Tiny sensors in each machine can monitor them for various factors such as heat, usage, vibration, pressure, overclocking, and electricity consumption. The data gathered is fed to a statistical model that can predict when machines will break down. Clients schedule service requests well before any event occurs. Tetra Pak operates over 5000 such connected filling machines lines globally.
With real-time data and the right analytics teams, businesses can cut maintenance by 20 to 50%, boost equipment uptime by 10 to 20%, and reduce maintenance costs by up to 10%. Controlling these costs can support differentiation. By allowing for greater awareness and control, predictive maintenance ensures better product quality, strongly impacting customer satisfaction and brand differentiation.
Analyzing the data and optimizing the predictive algorithms requires having access to hundreds of experienced data scientists and researchers on tap, which is a tall order for most companies today. Instead, smart companies turn to open innovation challenges. An open talent approach can revitalise their innovative streak and solve grand problems at scale.
A balanced predictive maintenance strategy allows for fast identification of issues and precise diagnoses. By simplifying the monitoring process, companies can reduce troubleshooting time and repair inventory quickly to increase the asset’s lifespan. Predictive maintenance lets companies order parts ahead of time. Maintenance can be scheduled effectively without incurring overtime costs. Planning repairs around downtime activities help keep the just-in-time inventory and supply chain management moving smoothly.