Andrew Ball, professor of Diagnostic Engineering at the University of Huddersfield, has warned against claims that the future of predictive machine maintenance should be led solely by data science.
Professor Ball believes that predictive maintenance should be driven by both data science and engineering expertise.
The professor was speaking at the latest COMADEM (Condition Monitoring and Diagnostic Engineering Management) conference, held at the University of Huddersfield.
In his presentation, Professor Ball noted that the separate techniques of detecting, diagnosing and assessing machine faults, with related prognosis, required the context offered by engineering expertise.
“I have attended conferences recently where speakers have talked about purely data-driven approaches to predictive maintenance, with no concept of what engineering really needs,” he said.
He continued by pointing out that data-driven methods are good at identifying patterns and anomalies in complex data sets, pointing diagnosis in a particular direction.
But this is just one element of predictive maintenance and while data science can be of significant assistance, it should be to the exclusion of engineering expertise.
Professor Ball was speaking at the 32nd COMADEM conference. Under the theme of Digital Enabled Asset Management, papers for the conference were submitted from around the world, including the US, Australia and China.