The aviation industry is grasping for opportunities to reduce costs. Big data has been making headlines in several industries, promising to revolutionize the way in which businesses are able to make decisions. One of the sectors slated to benefit from the use of big data, and associated analytics, is the aviation industry. As new aircraft generate more in-flight data compared to older ones, innovative analysis methods summarised by Big Data Analytics enable the processing of large amounts of data in short amount of time. Last studies show a reduction of maintenance budgets by 30 to 40% if a proper implementation is undertaken. As a result, predictive analytics of flight recorded data is an exciting and promising field of aviation that airliners are starting to develop. However, data sensitivity and security are some added complications that must be overcome through strong bonds between the MRO, flight operations and engineering departments to ensure all the data is employed. Unlocking the valuable information within this data is referred to as Data Mining. Although used in several other industries, the use of these tools in the analysis of aircraft data is relatively new and upcoming in recent years (Canaday, 2013).
The increase in data
It is a fact that nowadays aircraft generate more data than ever. Currently, around 2 million of terabytes of data are generated every year by the global fleet, through the Flight Data Recorder and Aircraft Health Monitoring. By 2026 this may have grown to 98 million of terabytes per year as shown in below figure. In the last decade several factors have caused a huge increase in data. First, the digitization of information keeps on progressing, but also costs of sensors, data storage and data communication has significantly dropped over the years. Finally, the velocity of the incoming data has increased enormously, caused by the advancing information technologies which make it easier to generate data (van Kempen & van Eijk, 2014; Chen, Mao & Liu, 2014). At the same time, these huge amounts of data need to be explored to discover meaningful and useful information. This makes manual analysis impractical (Mosaddar & Shojaie, 2013). Datamining presents an opportunity to increase the rate at which the volume of data can be turned into useful information (Bastos, Lopes, & Pires, 2014).
Figure 1: Data Generated by global Fleet - Fleet & MRO Forecast by Oliver Wyman
The main question you might ask yourself: 'Where are we currently standing: is our airline ready for predictive maintenance?'
Schedule a free consultation session about predictive maintenance with one of our consultants to answer this question.
Reduce costs thru Predictive Maintenance
A subject within Big Data Analytics and Data Mining is predictive maintenance. Predictive maintenance is recognized by 66% of the airlines as one of the most prominent new technologies to have entered the market by 2020. Also, Big Data Analytics is being used by 54% of the airlines to enhance Maintenance Repair and Overhaul (MRO) systems, and almost 92% plan to use their fleet data to improve health monitoring and predictive MRO (Canaday, 2015). Until now, it has mainly been done by Original Equipment Manufacturers (OEM) and OEM shops, as operators and MRO’s want to see more proven track records in the technology and processes.
Figure 2: Importance of Data Mining for predictive maintenance
The difference between preventive and predictive maintenance
Figure 3: Difference between predictive and preventive maintenance
Is useful when a strong correlation between equipment age and failure rate exists. For example, when abrasive, erosive, corrosive wear takes place or when the material properties change due to fatigue. In this case, the individual components and equipment probability of failure can be determined statistically, and the replacement of components is scheduled at a certain number of cycles.
Can be described as “the intelligent way to maximize machine availability”. With the right information in the right time it is possible to determine the condition of in-service equipment in order to predict when maintenance should be performed. As a result, it is possible to conveniently schedule corrective maintenance actions, preventing equipment failure. Compared with preventive and reactive maintenance tasks are performed when warranted, right on time.
The adoption of Predictive Maintenance is growing. Nowadays AHM is necessary, especially regarding Engine Condition Monitoring, followed by airframe maintenance and component maintenance. However, competitive efforts by OEMs to claim ownership of data generated by aircraft systems Carriers must preserve full rights to their data, including the ability to share it with maintenance providers