Improve Reliability in Chemical Manufacturing with Predictive Asset Analytics
This is the final blog in a four-part series on digital transformation in the chemical industry. If you missed them, check out the first blog “Chemical Companies Deliver Outstanding Customer Experiences Through Digital Transformation”, the second post “5 Benefits of Digital Transformation in Chemical Manufacturing” and last week’s post “Augmented Reality in the Chemical Industry”.
If you’re like most chemical companies you probably have a good bit of capital tied up in your MRO inventory. The question is whether or not you really need to keep that capital tied up in that inventory or not. A recent ARC reliability study showed that 82 percent of all assets have a random failure pattern. That means that even with all of the preventive maintenance, careful logging and reporting of service records, the very best maintenance, reliability and operations personnel simply can’t account for those random failures that no one sees coming. Or can they?
See the benefits predictive asset analytics offers for chemical companies in our infographic.
The Benefits of Predictive Maintenance
Typical maintenance strategies start with preventive maintenance based on calendar time or usage. But how does one account for things like manufacturing defects in equipment or simple human error in maintenance procedures? That’s where shadow sensing technology, machine learning, big data and predictive analytics come into play. These new technology resources are allowing companies to shift to a new, IIoT-enabled proactive maintenance solution. These solutions replace conventional reactive maintenance strategies with a strategy that can lower MRO inventory costs. Think of it as seeing what’s going to fail weeks or months before it does. That means you can perform maintenance at the optimal time, instead of replacing equipment that still has a useful operational life ahead of it, or scrambling after a failure occurs.
More accurate and efficient automated data collection enabled through IIoT sensor and information technology drastically expands the number and variety of environmental and operational parameters that maintenance and operations technicians can use to keep their systems running. The downside is that this big data can be cumbersome to actually sort through and manage. That’s where machine learning and predictive asset analytics systems come into play.
User-Friendly Predictive Asset Analytics
The problem with most solutions available today is that implementing them is more complex than the marketing we see. These solutions almost always require a data scientist to accurately train models to help detect operational anomalies. And that’s where things quickly get out of control in terms of complexity and cost. Like all new technology, there’s usually a guru behind the scenes to turn the dials and read the gauges to actually squeeze new value out of the solution. In the long run, that kind of defeats our purpose of reducing maintenance and operations related costs like MRO inventory.
But what if there was a commercially available, off-the-shelf solution that could handle all of that data science for you? What if it did all of the complex model building with minimal human input using data you’ve already been collecting? That’s where predictive asset analytics from Schneider Electric comes into play. Chemical companies have already seen dramatic impact with predictive asset analytics. BASF implemented predictive maintenance as part of their digital transformation initiative, improving their asset performance, reliability and utilization while increasing production efficiency.