Say goodbye to downtimes thanks to predictive maintenance
What if there were no more machine downtimes? What if problems were fixed before they happened? With predictive maintenance, that's possible. Alberto Catani, Project Manager Lean Production, Francisco Rosas, Project Sponsor, and Markus Klien, Global Maintenance Manager at thyssenkrupp's steering specialist, worked in an international project team to reduce downtimes with predictive maintenance.
The topic of predictive maintenance came up at the thyssenkrupp Presta plant in Puebla, Mexico, due to an urgent machine failure. The motors of a group of machines for tube forming required maintenance on short notice. For production, this meant unexpected shutdowns during which production could not take place. The result: high costs and a lot of lost time. A solution was needed to avoid unexpected downtimes in the future and to prevent it through predictive maintenance.
Together with the international colleagues Tod Turner, Global Head of Maintenance at thyssenkrupp Dynamic Components USA, and Markus Klien, Francisco Rosas and Alberto Catani developed a predictive maintenance concept for the site in Mexico. "In the worst case, the failure of one machine means a whiplash effect on the other highly interlinked follow-up processes," says maintenance expert Francisco Rosas. "With the implementation of predictive maintenance, we were able to reduce the downtime of the affected group of machines by at least 50 percent."
With the help of motor vibration measurements, the experts at thyssenkrupp Presta now monitor the actual state of the machines and can thus act at an early stage if a potential defect appears.
Reduce downtime with predictive maintenance
But how does predictive maintenance actually work? With predictive maintenance, the actual state of plants is monitored 24/7 in real time. This data transparency helps to identify problems before they occur. Through planned maintenance intervals, the causes of potential damage and the need for repair can thus be fixed without having to stop production plants for a longer period of time.
The goal of the thyssenkrupp Presta colleagues in Mexico was to improve process stability with the help of predictive maintenance. For the team, that meant engines with little to no downtime. "The vibrations of the 15 motors in the five plants of our machine group were measured, characterized and modeled with a measurable system to build a mathematical neural network algorithm that can predict breakdowns," explains project manager Alberto Catani.
In addition to the motors of the five plants, all head plants of the production at the plant in Puebla were also included in the predictive maintenance program. "With the help of Francisco Rosas at the plant, we were able to gain the first experience in predictive maintenance of head plants at the Eschen site," Markus Klien tells us.
Challenges and opportunities of predictive maintenance
The most important requirement for such predictive planning is measuring the technical and mechanical processes within the production chain. "Today's development of the necessary sensor technology, machine learning algorithms and predictive maintenance have increased fast in the last ten years due to the pressure of Industry 4.0," says maintenance expert Rosas. Accurate measurement requires vibration measurement devices for rotating machine elements, sensing elements, infrared thermometers for heat measurement and ultrasonic measurement devices.
"In addition to the technical requirements for predictive maintenance, it is above all the know-how within the company and its long-term safeguarding that is important for the success of the measure," explains Alberto Catani. Because a high fluctuation of employees always involves the risk that well-trained colleagues and their knowledge will go to the competition.
"Ultimately, together with the project management, we decided to train our qualified employees," says Francisco Rosas. In this way, the team is investing in the future. After all, predictive maintenance brings clear economic benefits and enables, for example, better planning of employee capacities, as unexpected maintenance and repairs can be significantly reduced with predictive measurements.
You can find more insights into the digitalization of our processes at thyssenkrupp in our stories.