APM, AI, Prescriptive maintenance, prescriptive analytics

How Does AspenTech APM Address Limitations of FMEA?

Asset Performance Management Solutions

6月 16, 2021

Please see part one of this two-part series Understanding the Impact of FMEA.

In part two, we discuss FMEA as a strategy component for optimizing asset performance and availability. A distinctive, contemporary Asset Performance Management (APM) approach, AspenTech offers its three layers of defense strategy. Our layered asset protection approach is practically similar to safety and cybersecurity tactics, incorporating risk assessments, guidance and smart FMEA.

 

  1. Preventive: includes regular service such as oil and filter changes, and routine manual inspections; typically, already in place to address wear-based failure modes. 
  2. Predictive: leverages sensor data to identify initial asset performance degradation patterns, that if unattended, will lead to failure. 
  3. Operational Risk: assures availability and full performance of all assets by providing total system-based risk/cost analysis, and protection against losses in quality, yield and waste. 

 

Preventive is what you likely already have in place with a systemized approach for scheduled maintenance tasks aimed at regular required service and inspection. However, predictive presents an opportunity to optimize maintenance efficiency, and operational risk moves the needle; assuring all equipment is available and performing its full intended duty. Each layer can also incorporate prescriptive guidance techniques providing clear, explicit instructions on what to do when errant conditions occur. 

Aspen Mtell® becomes the first line of defense, the primary shield. It actually performs preventive maintenance inspection tasks; not once-a-week, or once-a-month, but every few minutes with weeks-and-months’ notice of degradation and impending failure. Mtell allows reductions in non-mandatory inspections, resulting in large labor expense savings. 

How does Aspen Mtell help with FMEA? First, a Failure Agent carries a specific failure signature that matches an exact root cause and failure mode. It’s measured and accurate, rather than guesswork. A Failure Agent can actually carry the actual FMEA result that matches the exact failure mode signature in the data patterns. Failure Agents can also reveal the exact prescriptive guidance on what to do to avoid the damage, make process adjustments, or maintenance intervention with minimal production interruptions. 

Second, an Aspen Mtell Anomaly Agent detects abnormal conditions; not relying on estimating techniques but using AI-enabled exact pattern recognition. assuring detected deviations are real and not false. Yet in any system for anomalies, a human must always intervene to decide the anomaly type and what to do. With Aspen Mtell. it is quick and straightforward to decide if it’s a mechanical or a process issue. However, consultants and end users are pushing FMEA libraries to improve the investigation process with precise determination of anomaly failure modes. But FMEA cannot be as helpful as desired, as it’s not precise and is limited to later-stage mechanical degradation issues. FMEA does not determine early-stage process behavior that causes damage, but only assists when it’s too late, when obvious symptoms of actual damage appear. 

However, AspenTech is providing additional AI methods to significantly automate anomaly investigation. The Anomaly Agent could determine the difference between a mechanical and process issue. So, instead of waiting for the FMEA library to determine when damage is done, an Anomaly Agent can dispatch explicit instructions for process changes to stop damage-causing process behavior. When mechanical conditions cause the anomaly, the FMEA library will be automatically invoked. The smarts in the Agent indicate the highest contributing sensors and an automated map into the FMEA library indicate the potential causes to the user. For anomalies, the user always has to decide the cause. 

By detecting minor degradation and dispatching prescriptive advice, Aspen Mtell Agents can prevent small problems from becoming major damage and catastrophic failures. For example, detecting a lubrication issue where changing the lube oil and filter avoids more significant bearing failures. 

But, there’s more to APM than FMEA and reliability alone. AspenTech presents the operational risk layer to comprehensively protect against other forms of losses in asset performance besides degradation and failure. 

 

  • Aspen ProMV™ detects and corrects yield and quality.
  • Aspen Fidelis™ assesses risk and cost of all asset decisions from a total system performance basis.
  • Aspen Event Analytics™ (AEA) enables operators to quickly assess and correct errant process conditions that if uncorrected may lead to further performance issues.
  • APM Insights defines clear, consistent work processes presenting data from anywhere in views and dashboards for any users. Users can see things the same way and work them the same way to assure best possible outcomes.

 

As one example, Aspen Fidelis can extend the FEMECA criticality assessment (or the alternative reliability probability numbers) for total risk assessment for all operational circumstances. Aspen Fidelis’ risk protection can cover a single asset and its performance in relation to other assets, multiple assets and the complete plant unit, with an understanding of how myriad conditions affect the manufacturing process—including weather, failures, spare parts, feedstock and product delivery logistics, and intermediate storage. The criticality and risk assessment extends beyond a mechanical issue on a single asset. 

The sum of the preventive, predictive and operational risk layers in a layered defense strategy provides more complete asset performance management and risk protection — and does it easier, cheaper and faster.

 

Visit our APM solutions page to see if it’s right for your business.

 

 

 

 

 

 

 

 

There was a problem storing your subscription

Leave A Comment