AVEVA Predictive Analytics

Predictive maintenance can generate a 10-20% reduction in OPEX. The AVEVA Predictive Analytics solution can provide early alerts and diagnostics of equipment issues days, weeks, or months before they fail. AVEVA Predictive Analytics helps asset-intensive organizations reduce equipment downtime, increase reliability, and improve performance, while reducing operating and maintenance expenses.

Organizations must go beyond calendar and usage-based maintenance to reduce or eliminate downtime. Excessive or insufficient asset maintenance greatly affects profitability. It also has a negative impact on sustainability.
AI-based predictive maintenance directly impacts reliability, safety, compliance, and profitability in operations.

AVEVA Predictive Analytics integrates with a wide variety of historical data management, control, and monitoring solution systems and can be deployed locally or in the cloud. The system is very scalable and can be used to monitor a single piece of equipment, a specific plant, or hundreds of remote assets across multiple centers.

The equipment that can be monitored includes:

An anomaly was detected. What are the next steps?

STEP 1: Which sensors are contributing to the anomaly?
STEP 2: What are the potential failures and related inspections?
STEP 3: What is the level of urgency before the breaking point?

The Main Advantages of AVEVA Predictive Analytics:

1. Reduce unscheduled downtime.
2. Prevent equipment failures.
3. Reduce maintenance costs.
4. Increase asset utilization.
5. Extend the useful life of equipment.
6. Identify underperforming assets
7. Improve safety.

Native Integration with AVEVA PI System

There is no AVEVA PI System extension that is more natural than AVEVA Predictive Analytics:

• Visibility for more people, through content integration into PI Vision.
• Predictive analysis results integrated into PI System for contextualized information.
• Integrated with PI Asset Framework, which enables more efficient model building.

Intuitive Model Building

Artificial intelligence and machine learning without code. It is not necessary to be a data scientist or software developer to create, validate, and implement predictive models with AVEVA Predictive Analytics. The software uses an intuitive, code-free user interface to develop predictive models for your equipment.

Alerts and Notifications

Users can set alert thresholds to notify if the deviation between actual and predicted values exceeds the allowed limits. Alert management is done through a powerful and easy web application.

Failure Diagnosis

• Visualization and representation of failure diagnostics, including failure trees for more detailed information.
• Probability in failure modes.
• Corrective actions with prescriptive analysis.

Predictive Analysis

Users have the ability to view raw model data and model results, compare the performance of similar assets of the same type, and view the effects of alerts. The tool is equipped with error diagnostic capabilities to help the user determine the cause of the identified anomaly and how to avoid it in the future:

• Global anomaly trend.
• Trend of deviations of individual sensors.
• Contribution of each sensor to the anomaly.
• Ranking of potential failures.
• Failure match trend.
• Prescriptive guidance for corrective actions.
• Prediction of time to failure.
• Tracking of cases from the start of the alert to corrective action.

Prediction of Time to Failure

As part of the workflow, AVEVA Predictive Analytics allows the end user to predict the time of failure to obtain an indication of the level of urgency associated with correcting the detected anomaly. Thus, a user can:

• Determine the risk level of a team and the urgency to trigger predictive alerts.
• Calculate the repair time or replacement under current operating conditions.

Automated Model Construction

Autonomously deploy new predictive models for equipment of the same type in a single action. Save valuable resources, reduce errors and ensure consistency:

• Minimize manual work.
• Template creation.
• Automatic cleaning of training data.
• Integration to AVEVA PI Asset Framework or existing historian.

Case Management

Case management integrates learnings from past anomalies with user activities / comments. This knowledge capture is of immense value in today’s challenges with an aging workforce. Through case management you can:

• View predictive trends of cases.
• Make better and faster decisions with greater access to information.
• Highlight relevant cases when investigating failure diagnoses for anomalies.
• Integrate learnings from past anomalies with user activities / comments.
• Capture knowledge and best practices.
• Track actions (who, what, when).

Predictive Monitoring at Scale

Efficiency, quality, scalability, data analysis, maintenance and collaboration. These advantages can help achieve better results, optimize operations and deliver a superior experience to end users.

• AI and Machine Learning algorithms are the easy part.
• Operationalizing AI at scale is the difference between success and failure.

Bring your Own Algorithm

It is now possible to include custom algorithms in AVEVA Predictive Analytics and add value to investments already made.
Data scientists can include specific algorithms in the predictive maintenance cycle of AVEVA Predictive Analytics using Python or similar languages.

Custom algorithms can be complemented with integrated model templates, data cleansing functions, alert workflow, fault diagnosis, prescriptive actions, forecasting, case library, and integration with AVEVA PI System to monitor all analyses in a single application.