Reducing factory equipment downtime with AI-based predictive maintenance

[mroche1]

Asset management with AI-based predictive maintenance

With more than 46 million circuit breakers manufactured every year, unplanned downtime at the Schneider Electric factory in Plovdiv, Bulgaria could translate into hundreds and thousands of customers in Europe not getting essential power equipment.

At the Plovdiv plant, a critical part of the manufacturing process involves using Trumpf marking lasers. These machines add traceable information and labels for components used in the circuit breakers.

"Trumpf lasers are one of the best on the market regarding efficiency," says Georgi Tronchev, Head of Innovation Hub at the Schneider Electric factory in Plovdiv.

If these lasers stop working due to an unplanned outage, the entire operation at the Plovdiv plant can grind to a halt.

Worldwide, the number of Trumpf machines installed at Schneider Electric factories exceeds 200. These high-performing machines have a highly disciplined maintenance routine. Yet, these machines experience about three unexpected downtimes every year, halting production.

"Unplanned shutdowns are part of every production," says Tronchev. "But we are always trying to avoid them."

A pilot was launched to identify a solution for the unplanned downtime problem. Additional aims of the pilot include reducing the cost of repair and maintenance and increasing the lifetime of the laser machines. The factory in Plovdiv, which has 27 Trumpf lasers, was one of the first vital sites selected for the pilot program.

Digitizing laser machines and adding PdM via “Attention App”

During the pilot, 8 of the 27 Trumpf lasers at the Plovdiv site were 100% digitized with EcoStruxure Machine Advisor and an AI-based predictive maintenance solution leveraging machine learning for anomaly detection.

The solution combines EcoStruxure Machine Advisor with the complementary Attention App, an application that provides relevant and precise information on the health of equipment and machines. The data is made available to end customers in the manufacturing sector for decision-making.

 

Thanks to data analytics, the solution provides robust anomaly detection, leading to increased uptime and cost savings. The application can quickly process any type of data and allows a machine builder to scale up to predictive maintenance.

Additional benefits include faster and more efficient maintenance and more reliable monitoring and troubleshooting.

With Attention App, anomaly detection can be sent right to your phone and services can be planned.With Attention App, anomaly detection can be sent right to your phone and services can be planned.

"The biggest advantage for us is that data are pre-analyzed and sent in a comprehensive way. In this way, we are capable of making informed decisions," says Tronchev.

By the end of the year, the Plovdiv plant plans to connect all of their Trumpf lasers to the solution.

A robust PdM solution via EcoStruxure openness 

"Thanks to the openness of EcoStruxure, especially Machine Advisor, we managed to combine the right solution with our own existing architecture," says Tronchev.

The solution is planned to go on to other factories with similar equipment―the full project scope is 200 machines across 18 sites.

 EcoStruxure Machine Advisor's dashboards paint a clear picture of asset health.EcoStruxure Machine Advisor's dashboards paint a clear picture of asset health.

The Exchange platform, which allows an ecosystem of partners, customers, and end-users to have direct access to dedicated applications, helps accelerate innovation on the problems that industrial plants are faced with.

For example, a community for data center operators on SE Exchange empowers businesses to ask questions and find answers to significant manufacturing challenges, as well as collaborate on solutions that help extend the life of machines.

So, are you ready to help solve some of the most pressing challenges at manufacturing sites? Register on Schneider Electric Exchange today.