AI-Driven vibration monitoring unveiled

Automation

At the recent “The Maintenance Days”, organized by the National Organization for Swedish Maintenance, one notable case study showcased by David Svahn, a vibration data engineer from Södra Cell, and Stefan Lagerqvist, COO and co-founder of Viking Analytics, highlighted the transformative power of AI and smart monitoring technologies in optimizing maintenance strategies.

Transition to Wireless Sensors

Södra Cell is moving away from traditional hand-held vibration measurement instruments in favour of wireless sensors. Currently, 3,000 sensors are installed, with a goal to complete the transition by 2027–2028. These sensors offer continuous monitoring, eliminating the need for manual rounds.

Despite the advances, the system generates approximately 1,100 alarms weekly, handled by only two engineers. Alarm thresholds are often determined based on intuition and require frequent updates. While existing tools like data filtering offer some relief, the need for more robust solutions is evident.

Role of Viking Analytics and AI Tools

Viking Analytics has developed Multiviz, an AI-powered tool aimed at revolutionizing vibration analysis. Currently tested on 5% of Södra Cell’s sensors, the tool analyses data across thousands of machines without requiring:

•             Machine-specific details

•             Pre-set manual alarm thresholds

•             Machine type classifications

Initial results reveal a significant reduction in workload, saving engineers 2–3 hours of analysis daily with just 5% coverage. Full-scale implementation could result in even greater time and cost efficiencies.

Case Study: Predictive maintenance in action

David Svahn illustrated the potential of predictive maintenance through a real-world example involving a screw conveyor in a washing press. When vibration data flagged a warning, a manual inspection revealed a damaged belt. The issue was resolved during a scheduled maintenance stop, avoiding an emergency shutdown.

The effectiveness of the repair was subsequently validated using both Multiviz and Södra Cell’s traditional tool, CondMaster, showcasing the complementary strengths of advanced and conventional systems.

Benefits and Future Prospects

Smart sensors and AI tools like Multiviz allow early identification of issues, minimize manual labour, and significantly improve efficiency. While only 5% of sensors are currently integrated, the potential impact of expanding this to 100% highlights the value of digitalization and AI in industrial maintenance.

These innovations align with Södra Cell’s commitment to sustainability, operational excellence, and industry-leading practices.

Credit to Kjell Arne for the original article. Published with permission from Nordiska Projekt (Stordåhd Kommunikation AB).

Related Articles
Lessons in long-distance telemetry
Long-distance telemetry systems are central to the operation of critical infrastructure across sectors ranging from utilities and mining to energy and water treatment. However, while modern wireless...
HMS Networks launches new N-Tron managed industrial Ethernet switch platform
The next-generation N-Tron NT7000 Series delivers fast startup, rapid network recovery, and hardware-based precision timing—built for modern industrial networks where seconds matter.
IMI resolves severe valve erosion at Wildcat Point energy facility
IMI has resolved severe erosion issues affecting a high-pressure spray water valve at the Wildcat Point energy generation facility in Maryland, United States, improving reliability and reducing...