SBIR Phase II: A Manufacturing Monitoring System Using Sound Spectrograms and Artificial Intelligence

Maijker Corp., West Lafayette IN

Investigators

Abstract

The broader/commercial impact of this Small Business Innovation Research (SBIR) Phase II project lies in its sound-based AI sensing technology. This innovation is poised to significantly elevate the quality and productivity of our nation's manufacturing sector. Central to this mission is its ability to "listen" to the subtle languages of machinery, translating every hum and rattle into actionable insights. This ensures early detection of potential issues, minimizes downtime, and drives optimal performance. Collaborative tests with large manufacturing enterprises are set to showcase its potential in not only averting costly disruptions but also championing a culture of accurate diagnostics and continuous improvement. The societal implications stretch beyond factories. This technology could be embedded in vehicles, bestowing even age-old models with a modern-day protective insight, or in homes, warding off unexpected appliance mishaps. By unlocking the 'speech' of machines, this SBIR project propels us towards a future where our relationship with machinery becomes more intuitive and proactive. Ultimately, this project is a stride towards a seamless conversation between machines and humans, ensuring enhanced safety, efficiency, and an enriched quality of life for all Americans. This Small Business Innovation Research (SBIR) Phase II project aims to push the boundaries of machine condition monitoring by harnessing the underexplored potential of internal machine sounds. Merging Industrial IoT (IIoT) with advanced AI, the project crafts a system primed to provide real-time health and status updates for industrial machinery. The innovation's intellectual merit is in its pioneering method of capturing and analyzing internal machine sounds, paralleling techniques used in AI-based speech recognition systems. The project’s mission is to develop a "machine speech" recognition system that decodes the subtle intricacies of machine sounds for humans. This initiative will enhance our understanding of machine operations and introduce cutting-edge predictive maintenance systems, drawing from a largely untapped data source: internal sounds of machines. Mirroring the precision of a medical doctor using a stethoscope to assess human health and fueled by robust neural network models behind human speech recognition, this project represents the fusion of AI and manufacturing. Empowering workers to "tune in" to machine sounds assures a profound understanding of equipment processes and performance nuances. Diverging from traditional solutions that lean on vibration or current monitoring, this sound-centric approach promises a comprehensive view of machine health, marking a pivotal evolution in manufacturing. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.