Predictive maintenance is a valuable tool for any company that uses industrial equipment. However, there are still doubts as to whether it actually delivers measurable benefits.
To clear that suspicion, Machine design he recently spoke to Philipp Wallner, business manager at MathWorks. It answers some common questions that companies have when evaluating the return on investment of forecast maintenance, as well as addressing the misconceptions and benefits that are present. loop.
Machine design: Can you describe some of the misconceptions companies have regarding forecast maintenance?
Wallner: Predictive maintenance has been mistaken for a “black box” solution, where operational data from equipment is used to somehow predict the useful life of machines. However, this view goes beyond the importance of land knowledge, knowledge that is unique to a particular environment. It plays an important role in developing programs that detect and predict failures.
Companies and manufacturers who use predictive maintenance on product lines have many benefits. And companies that don’t use it can be at a competitive disadvantage. But the benefits of predictive maintenance are within the reach of any company willing to invest the resources to combine machine learning and ground knowledge.
A health check and maintenance apps developed at Mondi can identify equipment issues that may reduce downtime and maintenance costs.
MD: How can companies better connect machine learning and land knowledge, and shut down the data science and engineering communities?
Wallner: Now is a good time for companies to foster collaboration across the data science and engineering communities. Data scientists have traditionally been involved in maintaining a history-based prediction in mathematics. However, their land experience is often lacking in the engineering community.
There are tools to simplify this process and build collaboration, such as software simulation programs. These programs allow users to generate powerful predictive maintenance algorithms while ensuring that these algorithms require less field data to train. The programs can also help users who are not familiar with forecasting maintenance, allowing them to try out different ways to collect and train data using apps. For example, the packaging and packaging manufacturer Mondi uses such devices to develop a health check and predictive maintenance claims that identify potential equipment issues. And it only took a few months to set up and helps reduce downtime and maintenance costs.
Apps like this sorting learning app help users who are unfamiliar with forecasting maintenance, allowing them to try out different ways to collect and train data.
MD: How important are predictive data maintenance applications?
Wallner: Failure data needs to be collected for companies to best train predictive maintenance algorithms; however, this type of data is difficult to access as equipment does not often break down, and equipment is expensive to manufacture. deliberately simply failing to collect data. To address this issue, software tools help data generation failure using simulation models to determine how physical equipment might operate in the field under different test conditions.
Software simulation tools simplify the process for generating failure data, allowing them to strengthen predictive maintenance algorithms without the need for as much site data for “proper” training. These tools also allow operators to use different methods for data pre-preparation and training prediction models. Baker Hughes, for instance, they used software tools to develop pump health monitoring software that uses data analysis and in-depth learning for predictive maintenance. As a result, the company reduced equipment downtime costs by as much as 40% and reduced the need for additional on-site trucks.
MD: What trends in forecast maintenance do you expect for 2021 and beyond?
Wallner: Over the next few years, companies should see the rapidly growing computing power of peripheral computers and business controllers opens the doors to a new dimension of software on production systems. Forecast maintenance will be both on-site and near equipment and will use data across multiple factories and equipment from different vendors. Also, these AI-based algorithms will run on both real-time and non-real-time platforms and systems, depending on the requirements.
In addition, there will be more cloud platforms, which will be the most powerful users of forecast maintenance and feed data from devices around the world into the platform. Despite some remaining skepticism, companies should prepare for cloud-based forecasting maintenance, as the cloud allows manufacturers to collect data from multiple domains and train predictive maintenance algorithms.