FROM UNPLANNED DOWNTIME TO PLANNED INTERVENTIONS
Unilin Panels' vision towards production reliability
Unilin Panels, a part of the Unilin Group, is a leading producer of sustainable and high-quality wood based solutions for the construction and interior sector. From chipboard and MDF panels, HPL and melamine faced boards, Unilin Panels operates several production sites primarily located in Europe. Currently, Unilin Group functions as a subdivision part of Mohawk Industries, Inc., the global leader in floor coverings.
In the chipboard production process, wood particles, sourced from recycled wood, are compressed and bonded together using adhesive resins, subject to high pressure and temperature in industrial-scale facilities. Reliability management plays a critical role in maintaining consistent and high-quality chipboard output. Unforeseen machine issues disrupting the manufacturing process can lead to structural flaws and performance inconsistencies in the final products. Having an effective reliability and maintenance strategy is essential to mitigate such risks and ensure optimal product quality and reliability.
Pioneering sustainable chipboard, Unilin Panels is leading the way with an innovative reliability strategy, reducing environmental impact and delivering significantly more reliable, high-quality products. Partnering with VersaSense, Unilin Panels has deployed AI and IoT-based solutions across their manufacturing facilities at Bospan and Spano in order to transform downtime into insights and planned interventions.
Harsh manufacturing environment, heterogeneous machines, unplanned downtime
As part of its reliability strategy, Unilin Panels sought a solution to mitigate unplanned downtime. Unplanned downtime causes a significant financial impact and disrupts both upstream and downstream operations of the continuous facilities, while also affecting overall production quality.
To combat unexpected machine failures, it was crucial to implement continuous monitoring on a variety of machines throughout the production facility. This includes pumps, fans, electric motors, reducers, as well as redlers and elevators. The objective is to provide clear evaluations of a machine’s health that can be easily understood by all maintenance engineers, whether or not they possess specialised expertise in machinery.
With hundreds of unmonitored assets in this harsh manufacturing environment and the need for continuous, non-intrusive monitoring and analytics to provide scoring and guidance, Unilin Panels decided to engage with VersaSense to address the predictive maintenance challenge. In addition to offering 24/7 monitoring and machine health predictions, VersaSense delivered a seamless installation of intelligent wireless sensors, activated relevant visuals, reports, and predictions from day one, ensured compliance with corporate cybersecurity regulations, and integrated predictions and critical alerts with existing factory OT and IT systems.
Factory-wide network of self-organising wireless IoT sensors
To support Unilin’s reliability team in continuous asset monitoring and health predictions across the entire facility, VersaSense deployed a highly robust and secure wireless mesh network. This network comprises hundreds of battery-powered wireless sensors installed on various machines and dedicated to monitoring vibration and temperature around the clock.
All VersaSense sensors are fully equipped with secure over-the-air reconfiguration capabilities. This enables remote reconfiguration and updates of every software element on the sensor as needed. Thus, for example, new or specialized algorithms for vibration signal analysis or pattern recognition can be effortlessly deployed for specific machines whenever necessary.
24/7 continuous monitoring of machine health
The reliability team installed sensors on a range of machines, such as pumps, saws, compressors, fans, redlers, and conveyor belts. Each sensor was set up to regularly gather vibration and temperature data and analyze it locally. The collected vibration measurements undergo comprehensive analysis, including examination of time-domain metrics and full spectral analysis, to detect any anomalies attributed to machine wear.
Through an intensive focus on combining sensor and production data with and AI-driven predictions, the system autonomously uncovers machine regimes within the data, contributing to a comprehensive understanding of the machine itself.
Enhanced insights through AI-driven machine health assessments
The collected measurements are transmitted to the VersaSense Enterprise IoT platform, where AI-driven software performs trend analytics, anomaly detection, and pattern recognition of previously known failures. The platform not only presents users with a straightforward representation of machine health but also furnishes machine experts with the details necessary for further in-depth analysis.
The reliability team has the flexibility to include feedback on detected issues and maintenance operations that were performed, allowing the system to recognise patterns and retrain. Furthermore, the platform facilitates machine-to-machine comparisons, providing insights into behavioral patterns relative to similar machines.
Finally, a major requirement was to integrate with the existing in-house IT/OT systems, linking detected or evaluated events to the current production context. This interconnected approach enhances the overall understanding of the machinery’s performance in the broader manufacturing landscape.
Modern Reliability Management. Less unplanned downtime.
In the first few months following the installation of the initial sensors, VersaSense’s AI-assisted predictive maintenance solution enabled Unilin Panels to prevent several incidents and significantly transform their maintenance operations. Through continuous 24/7 monitoring of machines and analysis of vibration and temperature data captured every 5 minutes, Unilin Panels can seamlessly identify typical machine regimes and anomalies, providing insights into required maintenance actions.
This proactive approach has led to a substantial reduction in unplanned downtime. Being able to timely plan maintenance interventions and minimising surprise disruptions to production schedules saved tens of thousands of Euros not only in production time and quality, but also in major replacement costs by preventing equipment failures before they escalate to disaster.
Moreover, the cloud-based platform offered by VersaSense has played a important role in improving overall equipment effectiveness (OEE) at Unilin Panels. The trend analytics and anomaly detection functionalities provide a deeper understanding of machinery health, empowering the reliability team to make more informed decisions. Technical engineers receive daily health assessments and detailed insights into the evolution of machine performance over time. Additionally, the system enables them to evaluate the effectiveness of their maintenance actions.
In addition, the capacity to provide feedback on maintenance operations, retrain the system, and compare machines with each other further enhances Unilin Panels’ ability to optimise their production processes. This fosters a more resilient and efficient manufacturing environment.