I am looking for a paper or return on experience regarding the use of Machine Learning in predictive maintenance in the context of data center?
I'm biased here, but my colleague Edward Chuah has done some research on identifying and predicting these kinds of failures, though we have not taken full advantage of these predictions, yet, at TACC. Perhaps, also, there's not as much ML in this as you'd like to see, depending on your definition of ML.
We have worked on implementing machine learning in predictive maintenance. We use enabling technologies such as machine learning and data science to provide predictive maintenance solutions for OEM’s and dealers. This solution will transform and analyse the sensor data embedded in vehicles (cars, trucks, EVs), enabling our customers to deploy this solution in the manufacturing industry (aerospace, mechanical and automotive) to reduce downtime and costs.