The idea behind sensor, data and Condition Based Maintenance requires human intelligence and expertise. For most asset managers, it is already difficult to manage and, control data and select relevant information for analyses. Predictive maintenance is not only about health assessment by sensors but also detection of abnormalities before equipment breaks down. Sensors could provide flexibility to design maintenance schedules to mitigate the risk of unplanned stoppings. The overall maintenance efficiency can eventually be improved by the implementation of predictive maintenance.
The availability of sensors and technological advancement alone doesn’t enable de explorative research of data. Although the speedy data flow and collection of abundant data enhance the potential of analytics, the next step, the adoption of condition diagnosis is the next big leap step. Only analyzing the stored data helps to mitigate failure and uncover the hidden patterns. Therefore, management faces at least three major challenges of transformation from traditional to advanced maintenance.
- The size of data will exponentially increase and accelerate due to the comprehensive use of sensor networks.
- The transition from conventional database to non-relational database requiring an infrastructure and expertise to process, and handle structured as well as unstructured data.
- Handling and understanding the petabytes or even exabyte of data will become real challenge expediting real time data processing.
Due to the complexity of data, the current processing techniques could not meet the demands. Due to the enormous amount of data, a cohesive platform for structured and unstructured data becomes an essential element. Analyzing the unstructured data is the first priority in decision-making and prediction (Li, Bagheri, Goote, Hasan, & Hazard, 2013; Muhtaroglu, Demir, Obali, & Girgin, 2013; Wielki, 2013). However, not all the unstructured data can be beneficial to knowledge development and decision-making process. The data must be fit for purpose of maintenance policy selection. Expertise on the future use and performance are critical to interpret the sensory information for predictive maintenance. Furthermore, the selection and adoption of suitable sensors is also important. Although data may offer tremendous insight to the diagnostics and prognostics, the pay-off from predictive maintenance is only available with appropriate sensors selection and adoption.
Regarding the complexity, collaboration of industrial expertise and scholars must be involved to have sufficient breadth and depth of knowledge to make proper use of data for maintenance policies. Then and only then the promise of high level reliability for excellence and reduction of the frequency of corrective maintenance, increasing machine performance and enhancing overall production reliability will be reality.