Analytics Cut Maintenance Costs
EP Editorial Staff | March 1, 2021
Cloud-based predictive-analytics software provides enterprise-wide access to asset-performance data while potentially reducing unplanned downtime by 25%.
By Kim Custeau, AVEVA
The cost associated with unplanned downtime is a problem faced by companies across multiple sectors. According to market-data specialist Statista, New York, NY (statista.com), in 2020, 25% of respondents globally reported the average hourly downtime cost of their servers was between $301,000 and $400,000. This is a large price to pay for inefficiency, especially when there is software available to help this risk.
Predictive analytics reduce equipment inefficiencies, which are a large contributor to unplanned downtime. By increasing equipment effectiveness, companies have saved hundreds of millions of dollars in early-warning catches, with organizations experiencing a 30% reduction in maintenance costs, 25% improvement in unplanned downtime, and 25% gain in workforce efficiency.
Overall, predictive analytics allow a company to better understand its individual asset performance. Specifically, predictive analytics benefit operations by learning the normal operational behavior of an asset and providing early warning when that behavior is trending toward a potential problem. While this is useful in any maintenance center, it is particularly useful in an operational setting to help stakeholders determine the remaining useful life of production assets. As a result, companies minimize unplanned downtime, better plan their maintenance windows, tune their asset strategy to a “just-in-time” repair policy, and gain much better control of spare-parts inventory costs.
Predictive-analytics software develops a series of normal operational profiles for each specific piece of equipment. The profiles are compared with real-time operating data to detect subtle changes in system behavior that often indicate an asset is trending toward a potential failure.
An advanced-alert manager and email-notification system provide near-real-time updates of how well a plant or system is functioning. The alert system identifies which assets are experiencing anomalies, and possible faults. Prescriptive information related to the type of asset, its condition, and recommended steps to remediate anomalous behavior are included in the alert. Each alert can be further documented in a case-management system, contributing to an extensive asset-performance knowledge base.
Alerts and notifications communicate when a deviation between actual and predicted values exceeds allowed limits. Alerts can be managed by alert category, level, criticality, duration, and frequency. Each alert event is also directly linked to a graphical trend for that asset. That trend shows event data, threshold limits, and times when the values are in alarm. Relevant users and groups can receive notification in real time if an asset is in alert status through customizable email-notification capabilities.
The data in predictive-analytics software includes several advanced statistical and model-based comparison applications and business-intelligence tools that enable users to spend less time searching for potential problems. Users can view the raw training data, see results of the model, and compare the performance of similar assets. This diagnostic ability reduces the likelihood that an engineer will attribute abnormal operating conditions to the wrong variable.
Reliability, Scalability, And Performance
Deploying predictive analytics through a cloud-based system allows companies to readily and easily boost predictive- and prescriptive-maintenance strategies and monitor asset health and performance from anywhere. Similar to on-premises versions, predictive analytics in the cloud leverage AI (artificial intelligence) and machine learning to analyze historical data and automatically infer correlations between operations data and asset performance. This provides new insights that can be used to optimize operations and maintenance of assets.
By leveraging cloud-deployed predictive analytics, organizations can easily share their preventive-maintenance strategy between different locations and departments within an organization. This is accomplished by leveraging data in dashboards, process- and asset-centric visualization and alerting, enabling a single lens of truth, and enhancing collaboration. Typically, this cost-effective software-as-a-service solution is simple to set up, manage, and use without additional IT infrastructure or staffing.
Cloud-based predictive analytics offer several operational benefits. These include deploying information across an entire enterprise, minimizing the need for specialized IT resources, and providing complete and accurate analyses by leveraging combined data sources.
Solve Operational Challenges
Predictive analytics are needed now more than ever as companies undergo the fourth industrial revolution. The need is compounded by the current requirements for remote working options and flexible on-site scheduling. Industrial customers are also seeking solutions that solve the needs of today’s connected worker. Thus, information must be easily available remotely and securely to make critical decisions.
Predictive and smart analytics can play a key role in assisting the connected worker with seamless collaboration, access to timely information, and enhanced visualization and insights. Industrial companies need strong digital infrastructure to be successful during and after the pandemic and tools such as predictive-analytics software are helping to guide them forward. EP
Kim Custeau is the Global Asset Performance Management Lead at AVEVA, headquartered in Cambridge, UK (aveva.com). Custeau leads the strategy for industrial asset-performance-management solutions, which helps AVEVA customers improve asset reliability and performance to maximize return on capital investments and increase profitability.