Challenge
Wärtsilä wanted to look closely at audio data to see if it could be used as a new technology to improve predictions in operations and maintenance. We aimed to analyze various sound patterns and frequencies emitted from machinery to detect early signs of malfunction. By implementing this innovative approach, we hoped to enhance our predictive maintenance strategies, reduce downtime, and ultimately increase the operational efficiency and life-cycle of the engine. This initiative was part of our broader commitment to leveraging cutting-edge technology for continuous improvement and excellence in our services.
Solution
We solved the challenge by starting our collaboration with Emblica using real audio data captured from a pilot installation. Emblica’s expertise in audio and image data analytics was instrumental in helping us understand the potential of audio data capture. By leveraging Emblica’s advanced knowledge, we engaged in open discussions about the possibilities of modern AI systems. This led us to initiate the first experiments, which eventually developed into working solutions. While next-generation technologies played a crucial role, the close collaboration between the teams was the most significant factor in building a successful solution.
Collaboration
We collaborated by combining Wärtsilä’s domain knowledge of the engine combustion process and phenomena with Emblica’s expertise in audio data analysis. This collaboration involved demonstrations, data exploration, visualization of results, and developing business models around the technology. Wärtsilä provided the essential context of ‘what to identify’ in engine operations, while Emblica focused on ‘how’ to process, analyze, and interpret those signals using advanced AI and machine learning techniques. Continuous dialogue between the teams ensured that Emblica’s technical insights were aligned with Wärtsilä’s practical domain knowledge and business needs. Regular weekly meetings allowed us to improve the implementation and deepen our mutual understanding of the capabilities of audio data.
Impact on the energy transition
This collaboration impacts the energy transition by enabling Wärtsilä to harness cutting-edge audio data capture technologies, enhancing predictive capabilities in operations and maintenance. By integrating Emblica’s expertise, we can drive data-driven maintenance strategies and reduce risks associated with key performance indicators (KPIs), ultimately supporting more sustainable and efficient practices in the energy sector.
Niko Vettenranta, General Manager, Digital Products & Autonomous Operations at Wärtsilä
Lotta Koponen, CCO at Emblica