Over the past several years process plants have become more familiar with PdM solutions like SAM GUARD, using them alongside their existing diagnostics support. Although each of these approaches is valuable and powerful in and of itself, a much deeper understanding can be derived from combining diagnostics and PdM into a single plant maintenance system. 

This was the thinking behind SAMSON’s 2018 acquisition of Precognize and its SAM GUARD predictive maintenance (PdM) solution, which uses artificial intelligence (AI) and machine learning (ML) to interpret the masses of data produced by process plants and produce early alerts about impending failures. 

Since the acquisition, the company has been finding new ways to combine their respective ways of looking at plant processes, systems, and equipment. In the course of this journey, we discovered the benefits of merging our historic excellence in diagnostics, especially in valve diagnostics, with Precognize’s PdM capabilities. 

In this webinar, we discuss how to bridge the gap between predictive maintenance and diagnostics solutions. 

We conducted several polls and here are some of the results:

Do you believe there’s value in monitoring regular equipment, (i.e. valves, pumps, heat exchangers, condensers, small compressors)?

The Complementary Strengths of Diagnostics and PdM

The fundamental assumption that lies behind predictive maintenance solutions is that failures are inevitable; what matters is to catch them early, so you can keep downtime and costs of repair as low as possible. Every PdM solution is based on monitoring equipment and processes so you can spot the earliest possible signs of a failure, and act swiftly before they snowball into a crisis. 

On the other hand, diagnostics are based on preventive maintenance, which assumes you can prevent failures as long as you replace or run maintenance on a particular part within a specific, preset period of time. The role of diagnostics is to optimize that replacement/maintenance period, according to the real-life usage and conditions that affect each part. 

Why it Matters to Bring Diagnostics and Maintenance Together

When you combine both predictive maintenance and diagnostics, you can reach a new level of efficiency. PdM solutions can help plants adjust their maintenance windows not just according to the theoretical condition each part should be in but based on its actual state, as derived from PdM analytical monitoring. 

One example where this can make a real difference is during turnaround. Turnaround is hectic, and there are always some parts that you aren’t able to address. It’s crucial to know which issues must be dealt with, and which are less critical and can be safely left for any extra time available at the end. 

By combining our diagnostics expertise with PdM insights, we can now deliver a report that helps you prioritize which valves need maintenance during the next turnaround, which could do with some attention, and which can be considered “optional.” 

In the future, we see the possibility to replace today’s rhythm of preventive maintenance and periodic downtime with continuous, asynchronous maintenance that addresses each part on an at-need basis, instead of according to an externally-devised schedule, thereby maximizing productivity. 

Download Our Webinar: AI-based Predictive Maintenance as a Service

Combining Multiple Datasets from PdM and Diagnostics

Vast amounts of data are already being collected both at the plant level and the individual part level – i.e. internal diagnostic data. 

AI-based analytical monitoring solutions like SAM GUARD’s Precognize use historical data and a deep digital model that we create of the plant to crunch process data, identify anomalies and detect which part is likely to fail in near future. 

New smart parts, with their diagnostics, add additional datasets that bring another dimension to the analytics. For example, Samson offers smart valve positioners which collect hundreds, if not more, data points about the valve, its internal pressure, how many times it fluctuated, etc. 

When brought together, the process data collected by SAM GUARD provides valuable context to the diagnostic data about specific parts. The mapping of the entire plant that we do as part of the SAM GUARD implementation gives clear indicators of how each of the various elements interacts or impacts the other. This gives process and maintenance engineers a holistic view of everything that is happening in the plant. For example, if they suspect a problem due to an anomaly in the process data, they can inspect the diagnostic data to dig deeper, to validate or invalidate the 

Challenges in Bridging the Data Gap

There are four main challenges to combining these diverse datasets, which we discuss in-depth in the webinar. These include:

  1. Data extraction. The data may be stored in multiple locations such as the DCS, in the cloud, or the data historian. 
  2. Data integration. Data streams from sensors include various types of data, such as time series data, static data, flags, and others, and it all needs to be integrated. While this requires hard work, it’s not an insurmountable problem. 
  3. Analytical integration. Diagnostics data is devoid of context, so it needs to be integrated with the rest of the datasets to enable accurate analytics. This can be done at different levels, but the end goal is a single analytics system without multiple steps. 
  4. Operational integration. Because each plant uses parts from multiple vendors, each with their own diagnostics, the control room is likely to have to respond to multiple monitoring systems. SAM GUARD helps connect these dots. 

Bringing PdM and Diagnostics Together Improves Productivity

Combining PdM and diagnostics data in a smart way can contribute to improving plant productivity and uptime; there are still some challenges to work through and SAMSON is at the forefront of this. Watch the webinar to learn more about the benefits and challenges of combining PdM with diagnostics, as well as the ways that SAMSON’s SAM GUARD helps resolve them.