Predictive maintenance programs are taking center stage as more companies embrace digital transformation. The use of advanced technologies seeks to enhance the autonomy of physical assets and optimize their productivity. More companies are appreciating the role of data in improving existing maintenance practices.
Credit: Crystal Kwok / Unsplash
Predictive maintenance (PdM) relies on condition-monitoring to evaluate the asset usage cycles and identify defects that may result in asset failures. Early defect identification allows companies to proactively fix underlying equipment defects, preventing failures that result in costly damages or undesirable downtime. The benefits of a PdM program are immense. Companies can enjoy uninterrupted production, lower cost of asset ownership, better operational safety, and extended asset life cycle.
It takes time to establish, implement and streamline a PdM program. The entire process requires investment in additional data management resources, upskilling of maintenance personnel, and a culture shift across the company. One wrong move could prevent the company from meeting its maintenance goals. Once the company begins developing a predictive maintenance plan, it must establish criteria for measuring its effectiveness. It could be measuring:
- Mean time between asset failures
- Cost of maintaining assets
- Quantity of maintenance work
- Impact of the PdM program on the productivity of assets
- So, what steps should be adopted when establishing a predictive maintenance plan?
1. Audit facility to identify assets for the PdM program
Starting a predictive maintenance program is a process. Theoretically, everything is appealing and seems easy to implement. In reality, the program disrupts established norms. That means that companies must begin with a pilot program. They must validate that a PdM plan solves predominant maintenance issues. The pilot program allows them to experiment with different technologies and evaluate the practicality of available data. It also facilitates budget estimations for facility-wide PdM implementation.
Companies need to conduct asset audits and identify possible candidates for the pilot phase. Ideally, the pilot program has to be cost-effective and deliver insightful results within a short time. The following classes of assets are perfect choices for the pilot phase:
- Critical production assets
- Assets that fail frequently
- Hard to replace or source assets
- Production assets in hard to access locations
Companies can peruse previous maintenance records and original manufacturer manuals or consult experienced maintenance personnel to create a simple database and use it later to advance the PdM plan.
Bryan Christensen, CEO, Limble
The author Bryan Christiansen is the founder and CEO of Limble CMMS
2. Evaluate resource and technology requirements
The next step is evaluating resource requirements for starting and sustaining the predictive maintenance plan. Implementing a PdM plan involves a shift toward a data-intensive maintenance program. The company checks the compatibility of existing technologies and infrastructure with PdM utilities. At this stage, the company reorganizes its maintenance workforce. It considers advanced devices for the acquisition and processing of real-time asset performance data.
Since PdM leverages different technologies, companies may have to invest in new data handling infrastructure. If they outsource technological solutions, they engage several vendors before settling on the relevant PdM solutions. The company must undertake preliminary training to acquaint staff with the paradigm shift in maintenance. That way, the implementation process becomes smooth. Everyone knows how the new sensors, facilities and tools will enhance routine maintenance workflows.
When evaluating resource requirements, the company needs to devise standard procedures for continuously assessing the effectiveness of the selected PdM technologies and resources. Such evaluations allow the company to phase out ineffective solutions at the early stages of implementation.
3. Select appropriate condition monitoring techniques
A PdM strategy is an effective means for companies to preempt asset failures by equipping them with sensors that monitor their performances. It is the stage where the company chooses to monitor specific operating conditions like machine vibrations. It requires the installation of accelerometers on rotary or linear motion equipment. If the company wishes to monitor more than one performance variable, it requires more sensors and data handling equipment.
The company should select appropriate condition monitoring techniques depending on the goals of the PdM strategy. It evaluates frequently occurring failures to decide on the best fault detection technique. Generally, companies can utilize one or a combination of the following condition-monitoring technologies:
- Vibration analysis
- Acoustic testing
- Oil analysis
- Infrared thermography
- Ultrasonic testing
- Dynamic temperature and pressure analysis
Focus on a technique that solves the company’s most troubled assets. With time, the company gathers sufficient sensor data to develop and improve algorithms for predicting asset failures.
4. Optimize data management
With the relevant condition monitoring techniques working, the company begins optimizing its data management practices. The company builds a reliable database using the existing sensor data. The database acts as a preliminary inference point for training defect detection algorithms. It allows the company to establish standard operating procedures for the PdM plan.
Optimizing data management prepares the company to up-scale the predictive maintenance program and revise its maintenance goals based on verifiable results. The optimization involves selecting dedicated data management teams and equipment, which can be in-house or cloud-based, depending on the company’s asset size.
Final Remarks
A robust predictive maintenance plan complements the company’s aspirations to digitize operations and maintenance activities. Like any change management program, its implementation undergoes rigorous planning and continuous evaluation. The company needs a foolproof strategy that proactively addresses challenges occurring at the implementation stages.
For a successful implementation, the company has to start small, set realistic maintenance goals, invest in the right technologies, and focus on steady improvements.
- The author Bryan Christiansen is the founder and CEO of Limble CMMS. Limble is a modern, easy-to-use mobile CMMS software that takes the stress and chaos out of maintenance by helping managers organize, automate, and streamline their maintenance operations.
Back to Homepage
Back to Technology & Innovation