Data Science delivers 70% downtime reduction for a machine manufacturer
When in operation, machines emit signals across a broad spectrum of frequencies. Some of them happen to be in a frequency range that the operator’s ears can pick up as sounds, and even though that frequency range is small, the information gathered is sometimes invaluable: good operators, maintenance crews, or mechanics, can often tell what’s wrong and what is right by listening to the humming of the machine.
ei3 has developed a broad spectrum sensor solution that is able to capture a range of frequencies that go well beyond what the ear can hear. And, unlike the ear, the sensor is always on the machine, always listening. Together with clever machine learning delivered by our data science team, the result is a solution that can detect machine anomalies early, allowing for predictive maintenance before a part actually fails. We can this solution “ei3 Symphony”.
Recently we put this solution to test. Our client, a European machine builder, uses a mechanical cutting mechanism on a revolving cylinder as part of machine design. The blades, mounted on the rotating cylinder, make perfect cuts as long as the alignment between the rotating blade and its stationary counterpart is perfect. As it goes out of alignment, cut quality deteriorates, ultimately leading to machine downtime. Alignment or replacement is a tricky exercise that usually requires the visit to a factory service technician. Therefore this typically means longer, costly downtime.
ei3 Symphony, mounted in close proximity to the cutting blade, is able to detect the deterioration of the cutting action long before it affects the actual cutting result. Alerts can be created long before the machine suffers from downtime. This particular customer estimates that 70% of downtimes are successfully avoided by using our sensor solution.