In our last post, we looked at how we’ve applied what we’ve learned about predictive maintenance over the years to everything we do today. Here we’ll explore how all of this experience and understanding has made Senseye PdM the leading predictive maintenance product on the market.
Years of experience
As we’ve previously mentioned, Senseye PdM is the result of over 150 person-years of research and development time spent exclusively on predictive maintenance. Building it wouldn’t have been possible without the skills of various industry specialists, mechanical engineers, and condition monitoring experts, in addition to a team of leading data scientists. And with a combined 30 years working in the aerospace and defense industries - world-leaders in the cultures of safety, maintenance practices, and predictive maintenance technologies - our founders’ extensive knowledge and experience of predictive maintenance have proved invaluable, too.
This level of experience and heritage, combined with our state-of-the-art machine learning technology, is what makes Senseye PdM a truly unique solution - one that’s made even more effective by the deep domain support and consultancy we can offer.
Designed with users in mind
Senseye PdM is unique in that, unlike other predictive maintenance products, it’s designed for the maintenance engineers that use it. We know how busy these people are. That’s why it doesn’t need continuous manual reviews of the sensor data it uses. Neither does it require the development of custom models for every type of machine it monitors. Instead, it automatically creates high-fidelity models for each machine, with no need for human intervention.
This means it’s possible to apply predictive maintenance to every machine within a given plant, even those of lower criticality. Using a combination of AI, machine learning, statistical modeling, prognostic, and data mining, the system can automatically analyze machine data, cancelling out any environmental noise that might distort the results, to accurately establish a machine’s health and forecast its degradation.
The results of this analysis are then fed into Senseye’s Attention Engine, a proprietary algorithm which generates an Attention Index® for each machine. A sufficiently high Attention Index will cause the Attention Engine to generate a Case that will direct a maintenance engineer’s attention to the machine in question.
And, because we know that maintenance teams have little time for trawling through huge amounts of data to find what they’re looking for, Senseye PdM presents them with everything they need to know in a clear and easily understood format, allowing them to react with speed and accuracy.
Our experience has given us an almost instinctive understanding of just what PdM is, and how it should function. In short, we see Senseye PdM as a decision support system that helps maintenance professionals take care of their machines. And this view forms the basis of a philosophy which underpins everything we do.
Here, then, are the three guiding principles behind the design of the application and of the analytics on which its success depends.
1 - Guide attention
The primary aim of the application, and the analytics that support it, is to focus the attention of its users on the machines that need it.
2 - Focus on meaning
Analytics need good data, high-quality algorithms, and plentiful context to generate anything meaningful. We know, however, that context is limited in manufacturing environments, so we focus on making our analytics work as efficiently as they can with the minimum context available.
3 - Model the user
We supplement the limited context we have with additional context from users. By focusing the analytics on the user, we’re able to predict user interest as well as machine health, for example. And because we have access to the user, we can use their feedback to optimize our predictions.
Each of these three principles comes from a deep understanding of predictive maintenance, born out of years of direct experience. And that’s what makes Senseye’s approach to the PdM problem unique.
In our next and final post, we’ll consider what the future of Predictive Maintenance will look like. Until then, you can contact us to learn more about what goes on behind the scenes at Senseye, or download our white paper - Senseye in Depth: Why is Predictive Maintenance so hard? - for more detail.