Cooperations for your success, Part 1: Predictive maintenance with Statworx and EASY
As we all know, working in a team makes everything better. We at EASY understand this all too well from our cooperations with our partners. And that’s why we’ve entered into partnerships for product development, so we can unite multiple areas of expertise and offer our partners and customers a broad range of outstanding products and solutions that are in line with the times. One of these partnerships is our collaboration with Statworx.
Solutions for the industry of the future
At Statworx, everything revolves around predictive analytics, machine learning, and artificial intelligence. The company is a start-up from Frankfurt currently employing 15 specialists in the data science field. Working with EASY, the company wants to improve core processes within the enterprise, establish new business models, and use existing areas of potential within available data. One example of how they’re currently working with EASY is the issue of predictive maintenance, a field that deals with predicting machine failures and optimising maintenance intervals. Every industrial company already collects sensor data from their production processes. Statworx combines this data with algorithmic methods to derive mathematical models they can use to forecast machine and machine component malfunctions.
Reactive maintenance still the rule
Most industrial companies wait until something fails: It’s clear that this kind of solution isn’t exactly optimal. Production stops, and production downtimes are associated with lost time and costs. In addition, there’s no way to plan the company’s requirements for service technicians and resources if machine downtimes are left to “chance”. Predictive maintenance the next stage in maintenance management. In predictive maintenance, technicians inspect production machines at set intervals. Of course, this doesn’t mean the company won’t ever waste resources, time, or man hours. In some cases, machines might not even have been in danger of breaking down. The next step, which some companies are already taking, is rule-based maintenance. In rule-based maintenance, companies analyse specific service intervals and the timing of past machine failures, then try to derive heuristic processes from this data. They can then draw conclusions about when the company will need to hire service technicians. But this still isn’t the last word on the subject: predictive maintenance can do even more.
Analysing gigantic volumes of data
Predictive maintenance uses many different kinds of data the company already has available to it, including data in SAP systems: data on historic downtimes, past maintenance intervals, and sensor data collected during products. It combines this with external data like weather, environmental influences, supplier information, etc. These are truly real data cases, where hundreds and thousands of influential factors meet millions of data points and modules, linked to mathematical algorithms on downtimes and forecasts. It’s even possible to generate self-learning rules from available data and prognoses. Not only production companies can use this kind of data; it’s also a great tool for large industrial operations like power plants and oil production facilities.
Process2Design is a major plus
The automotive sector has been familiar with such processes for quite some time. When you drive to the workshop, a mechanic connects a device to your engine that can read out all kinds of data, like when and how long the driver hit the brakes, then link it to external data. All of this data then gets bundled into pre-programmed algorithms capable of assessing and weighing the key influential factors and creating forecasts for when components will fail. Now, let’s image we add EASY PCM Process2Design to our example: The solution combines forecasts from predictive analytic models with a SAP system interface, initiating specific processes and actions by company mechanics. We can transpose the example above onto all kinds of production processes and industrial systems: Welcome to the world of tomorrow.
Are you interested in predictive maintenance?
You can also read more in the 2nd part of our series, when we will be presenting our cooperation with iXLOG.