Using predictive analytics to move from reactive to predictive maintenance
More and more devices generate data. That offers a lot of insight but also many possibilities. If you save and analyze the data correctly, you can improve a lot of the efficiency of the devices around us.
If you know how things have behaved in the past, you can often make better predictions about the future. And that saves a lot of time and money.
The data comes in via sensors on the device and gives information about the state it is in and how it is used. That way you know on time when it is due for replacement or for maintenance.
An example. Via sensor data that comes from aircraft turbines, it can be read when it can best be offered for maintenance. We can predict the incoming data when engines start to show a major breakdown and, in some cases, also what type of breakdown.
The aim is to prevent major disruptions, because they cost a disproportionate amount of money compared to planned maintenance. The level of requirements is high due to the high safety standards.
Through continuous monitoring we have insight into what a turbine really does instead of taking the average wear and tear. In addition to unexpected early disruptions, we also see that an appliance will last longer than expected. Type of use and circumstances are very relevant to the differences we see. Because of this insight, we can plan much better which turbine should come in for scheduled maintenance instead of as a result of an incident (major failure).
VIQTOR DAVIS analyses the sensor data and develops prediction models for, among other things, predictive maintenance. Based on these models and artificial intelligence, we can support smarter processes and business concepts.