Founded in 2017, Viking Analytics is heavily involved in developing predictive maintenance and data analytics solutions, which is reflected in the MultiViz solution. We’re talking about a specialized digital tool that gives organizations the ability to monitor and understand machines with, as they describe it, ”unparalleled precision.” This AI-powered solution not only identifies machines that require attention, but also offers detailed insights for maintenance recommendations. Collaboration with experts enables continuous feedback, which in the guise of AI capabilities and ML (Machine Learning) allows the tool to ”learn” things and evolve. Consistently installed and implemented, MultiViz’s collaboration methodology is said to ensure reduced false alarms, enabling companies to focus on the right machines and measurements at the right time.
Benefits and future prospects
During the presentation at the Maintenance Days, David Svahn illustrated the potential of predictive maintenance through a real-life example involving a screw conveyor in a laundry press.
When the vibration data flagged a warning, a manual inspection revealed a damaged belt. The problem was resolved during a planned maintenance shutdown, which meant that an emergency shutdown could be avoided.
The effectiveness of the repair was then validated with both MultiViz and Södra Cell’s traditional tool, CondMaster, which in this case demonstrated the value of the complementary strengths of the combination of advanced and conventional systems.
Moreover, Lagerqvist claimed that smart sensors and AI tools such as MultiViz allow early identification of problems and can be of great industrial benefit in machine environments. They have the potential to minimize manual work and improve efficiency significantly.
At present, however, only 5% of the sensors in Södra Cell are integrated, which ultimately highlights the potential effect of expanding this to 100%.
Notably, Södra Cell 2021 installed a thousand wireless Airius vibration sensors with temperature measurement, whose measurement data is transmitted via the mobile network. Measurement data is stored and analyzed in the company’s Condmaster software, which is installed in Södra Cell’s network and where the web application Condmaster.NET is included for access to measurement data via mobile, tablet or other screen. Measurement data can be integrated into other IoT systems at Södra Cell through the REST API and UPC UA.
These innovations are in line with Södra Cell’s commitment to sustainability, operational excellence and industry-leading methods.
What can Viking Analytics MultiViz do?
Viking Analytics software enables vibration and production engineers to analyze data and, via integrated AI capabilities, provide effective recommendations. Features in the environment are the following – solutions that:
- Automatically identifies “suspicious” machines without relying on manual setup or fine-tuning of threshold-based alarms
- Automatically identifies operating modes and detects early machine behavior that indicates an impending failure
- Focuses on the machines and the measurements that require the attention of experts
- Collects feedback from experts to learn the most relevant patterns and further refine the AI
- Enables machine experts to build reliable, scalable and generalizable models for different machines, vibration sensors and process sensors
- Enhances labeled data to build truly generic and scalable models for different machines, machine types, vibration sensors and process sensors.
