Enterprise Asset Management (EAM) systems enable improvements in the management of asset maintenance and the application of planned maintenance procedures with the goal of reducing unexpected maintenance and unplanned downtime. The introduction of Predictive Maintenance (PdM) solutions build on this goal by introducing technology that allows an asset’s performance to be monitored, analyzed and evaluated in real-time or near real-time. Dashboards delivering these predictive analytics, enhanced by machine learning algorithms provide greater visibility into the asset’s status and performance. In addition, automatic anomaly detection notifications and integration with standard EAM systems enable the optimization of an organization’s resources to further minimize the cost of unexpected maintenance and unplanned downtime. PdM is therefore typically conceptualized one-dimensionally with a focus on greater savings to be generated from greater reductions in unplanned maintenance and unplanned downtime.
Manufacturing organizations seeking to determine the real Return on Investment (ROI) for a PdM solution should not overlook the operational and financial benefits to be realized by the positive effect on downstream processes. The full benefit of a well implemented PdM solution will be realized in ways far beyond the minimization of unplanned downtime. For example, consider the downstream benefits associated with:
• More accurate production scheduling
• Reduction of MRO and spare parts inventory
• Improved product quality
• Improved employee and environmental safety
To fully appreciate the impact of operational advantages on overall ROI of a PdM solution you must take a multi-dimensional view. Let’s look further into these considerations.
Accurate Production Scheduling: PdM solutions can lead to more accurate and optimized production scheduling through improved knowledge of an asset’s condition and availability for use, providing users with greater insight into the impacts to near-term scheduling expectations. For manufacturing operations, knowing the impending risk for unplanned downtime creates an opportunity for more strategic production scheduling. Below are a few examples of how production scheduling will be impacted by PdM:
1. Use compartmentalized production to avoid the risk of a wasted product. If the PdM solution uncovers an impending period of higher risk for an asset, production efforts can be halted prior to the high-risk phase. Once the risk period is mitigated, there will be a surplus of work in process (WIP), but the reward is avoidance of waste caused by well-made WIP components being ruined by the at-risk asset creating flaws.
2. Increase the accuracy of time allotted for completion of planned maintenance. We know that planned maintenance does not always go as planned. Discovery of additional issues requiring attention may lead to unplanned repair extending the downtime period longer than originally expected. PdM will provide very specific indications of where higher risks for problems exist prior to the start of planned maintenance so maintainers can plan accordingly and avoid surprises. The benefit to production scheduling is more confidence in the expected planned period of downtime.
3. Respond to increased risk with alternative product runs. Often, the same asset may produce a diverse line of products or products having different attributes such as color, size or shape. PdM can provide insight into the risk of problems specific to each different production mode. If one production mode signals an amplified risk for a specific problem, the schedule may be altered to deploy lower-risk products as a temporary alternative resulting is a more flexible and optimal production schedule.
Reduction of MRO and Spare Parts Inventory: Manufacturing organizations that can accurately predict demand, usage and vendor performance have an edge in optimizing MRO and Spare Parts Inventory. Unfortunately, this remains a challenging task for many warehouse management personnel. With the introduction of PdM, spare parts, especially critical components, become likely candidates for IoT and Predictive Analytics, leading organizations to a more thorough understanding of component performance. This understanding will support and improve the planning and budgeting process. Access to the age of the component, number of repairs, failure history, current condition and estimated life expectancy becomes critical to reducing inventory and accurately planning for replenishment and replacement. Predictive Maintenance provides a common platform for improved planning and forecasting of MRO components and replacement parts.
Improved Product Quality: Over the years, manufacturing companies have used technology, process, and people to reduce product quality issues identified on the shop floor. IoT has added an additional dimension to improving quality programs by providing a mechanism to collect data from the shop floor and blend it with other available data sources allowing for Artificial Intelligence (AI) driven solutions. Additionally, fog and edge software have enabled AI solutions to be on or close to the equipment that might be responsible for quality issues. So how does implementing a PdM solution benefit quality?
1. Quality problems are often directly related to the degradation of the production equipment causing the problem. With no additional investment in a Quality solution, a manufacturer will see fewer quality impediments simply from ensuring assets are maintained prior to failure. The machine learning models built into the predictive maintenance solution will identify assets entering a state statistically leading to failure and notify maintenance in this early stage of degradation. For example, if a manufacturer identifies increased risk early and prevents a piece of equipment from degrading further or unexpectedly failing, not only has the downtime of the asset been avoided, but the quality of the downstream product will not be negatively impacted.
2. For a further reduction in quality issues, consider implementing a quality solution that will include new machine learning models built specifically for the quality function, and include new data sources blended with the data collected by the PdM solution. Using the same data collected for PdM and complementing it with additional data sources like video or imaging, a manufacturing company can quickly identify anomalies or quality related issues. For example, your PdM solution predicts increased risk of failure on a particular assembly station, this failure prediction becomes a data input influencing a predictive Quality model, which then alerts users that products downline from the assembly station have an increased risk of quality issues.
Employee and Environmental Safety: Safety has long been a concern with manufacturing companies due to the obvious concern with the well-being of their employees and the surrounding community. There are also many additional hidden costs such as lost employee time, workers’ compensation settlements, increased insurance premiums, fines and penalties and even damage to the organization’s public image.
1. A PdM solution will immediately contribute to the avoidance of employee safety incidents by reducing the number of dangerous tasks required of employees. For example, certified personnel would no longer have to enter confined spaces to do routine inspections after placing technology such as sensors or cameras in those areas to collect data.
2. Implementing a PdM solution and acting on the insights and predictions it provides, will immediately reduce the risk of accidents related to unexpected equipment malfunction. Preventing certain critical equipment from failing or degrading dangerously will reduce or eliminate employee related injuries and environmentally damaging accidents.
3. Employee safety can be further enhanced by OSHA certified wearables. The data collected from wearable technology serves as a wellness monitor for specific types of injuries and will ultimately reduce work compensation claims. However, blending data from a PdM solution together with data collected from wearables should significantly reduce the risk of employee injury or death.
With traditional Enterprise Asset Management solutions, companies frequently continue to struggle in their attempts to significantly reduce unexpected maintenance and unplanned equipment downtime. Predictive Maintenance solutions can now provide the timely, actionable predictions required to optimize asset performance and extend asset lifecycles. However, the real ROI of a PdM solution includes many opportunities to improve downstream processes and gain operational advantages over your competitors.