What if we could embed expert-level intelligence into devices like smartphones, enabling built-in microphones to “see under the hood” of complex systems?
Josh Siegel, assistant professor in the Department of Computer Science and Engineering at Michigan State University, is exploring just that. By knowing when parts will fail, equipment operators – from automotive and robot manufacturers to infrastructure builders and managers – could increase efficiency, sustainability and resilience.
Siegel turns data gathered via a low-cost app into near real-time insights, providing days or weeks of lead time to address faults compared to conventional diagnostics. This allows users to:
- minimize warehousing space and cost because they know when they’ll need parts
- prevent engine or transmission failure thus keeping drivers safer, and
- eliminate downstream failures that cascade and multiply costs
Plus, the software’s accuracy improves over time and the app can be used with assets in aggregate. For example, fleet operators could reduce the number of vehicles needed by using predictive, rather than preventative, maintenance because downtime is predictable and unexpected failures are minimized.
To learn more about Siegel’s research, visit:
Improving Misfire Fault Diagnosis with Cascading Architectures via Acoustic Vehicle Characterization [Article]
Surveying Off-Board and Extravehicular Monitoring and Progress Towards Pervasive Diagnostics [Article]
Innovation Center, Josh Siegel profile page [Article]
MSU College of Engineering Marketing and Communications page