How to start with data analytics in aircraft maintenance although much information is still paper based?
In 2011, IATA embarked on the paperless maintenance initiative called ‘Paperless Operation Initiative (PAO)' by publishing a series of guidelines around digital “best practices” encompassing key areas: parts tracking, electronic maintenance records keeping, aircraft deliveries and redeliveries, electronic logbooks, eSignatures among others.
IATA setup a series of dedicated working groups tasked with sketching the digital roadmap templates for the identified areas. The idea behind the initiative was to target year 2020 as a soft deadline for the Aviation Ecosystem to go paperless.
Nine years on, progress has been made but not nearly to the extent that would do justice to the goals behind the initiatives set. While there has been a significant push to move away from paper to a paperless environment, a lot of the historic data needed for harnessing the capabilities of Data Analytics and Artificial Intelligence (AI) tools today are still in paper formats.
When Airlines and MROs undertake digital transformation journeys with the aim of using historic data for data analytics and predictive maintenance, it involves tapping into data sources both electronic and paper. An additional intangible layer exists in the shape and form of human experience.
Take, for instance, an MRO that has data pertaining to hundreds of heavy checks carried out over the course of their business: a lot of valuable data – linked to the type of aircraft / engine serviced, age of the aircraft / engine, history of Non Routines and associated spares & man-hours data – remains either in paper or archived systems. The potential to reuse this data for forecasting costs and Non Routines would offer significant advantages and savings / efficiencies in terms of quotation costs & profitability, and material & capacity planning.
Three steps to take before starting with an analytical project
Considering the above elements forces results in the fact that you first need to:
Digitize paper data through advanced OCR tools: Taking the example quoted above, work pack details in paper / scanned PDFs can be subject through OCR tools (bear in mind, most OCR tools may not recognize aviation standard forms and templates & would require heavy configuration that takes time) to extract, consolidate and index information by Aircraft Type, Age, Check Type and details pertaining to Man Hours, NRCs and spares & tools used.
Audit and verify data sources: Information residing in paper and electronic formats may not tally with one and other (e.g. Form 1 data versus what is stored in a system electronically). A process to compare the data, audit and verify the sources is mandatory before pushing the data in for any further analysis. Whilst in theory a very manual process, a lot of this can be semi-automated through Robotic Process Automation (RPA).
Eliminate data islands to create a single source of truth: After carrying out the above two processes, it is necessary to create a single data warehouse that maintains the audited and verified information. This, in essence, is the single source of truth that algorithms and analytical engines tap in to for running mathematical models and further analysis.
Airlines and MROs that combine capabilities of Humans and Analytical Insights will harness better customer relationships, experience enhanced productivity by making their data work for them.
Business benefits for airlines and MRO’s to use data analytics
Creating this source data base allows an Airline or MRO to define KPI’s that would, in turn, link to tangible and intangible business outcomes. To quote a few examples:
Effective utilization of check maintenance intervals by Aircraft / Fleet
Automated Reliability Reporting and trend monitoring
Base Maintenance predictive findings for a mid-life B737-800 that has primarily operated under harsh environments
Engine Shop Visit Build goals based on “factored” Remaining Useful Life in LLPs, EGT Margins & Fuel Burn Rate
Just In Time (JIT) spares planning based on demand forecasting including Vendor Performance Metrics, Rotable failure rate modeling
Each KPI would require a combination of various historic and dynamic data sources for accurate monitoring and trend forecasting. Each KPI is linked to a business outcome – process efficiency (e.g. improved dispatch reliability, fewer AOGs), cost reductions (e.g. higher profitability in MRO quotations through greater accuracy in NRC predictions), faster TAT or enhanced communication with supply chain vendors.
Once the Airline or MRO identify the KPI’s and expected business outcomes, the data fields within single source of truth data base are passed through mathematical models that resemble the nature of your operations.
What you can do right now
However, to make your data work for you, the first step is to work towards ensuring the creation of a single source of truth. This process, though cumbersome, can be semi-automated through tools such as advanced OCR, RPA and systems interfacing. With EASA publishing its first version of the Artificial Intelligence roadmap this February, it is a sign that there is openness towards an increased adoption of using a combination of Data Analytics and AI to realize organizational goals.
To put in a nutshell
Most organisations are wary of data challenges within their Airline or MRO, but they aren’t as familiar with the potential that unearthing and enriching their own data can offer in terms of valuable insights. With technology growing at a rate several folds faster than business, adopting process and tools that catalyse their journey towards enhanced benefits realisation requires a thorough assessment of where they are now, where they aim to be and build the digital roadmap. After all, analytical insights without the right data sets remain of no value.
How EXSYN can help
We partner with Airlines & MROs to help define digital roadmaps and build insights that they need to scale up efficiencies using a combination of our experience and their own data. We take an active role in training the customer resources through workshops on adopting state-of-the-art tools and processes as part of the transformation.
Having supported over 30 airlines and MROs spanning across a fleet of 1450+ aircraft, our products and services are tailored to meet your specific analytical goals.
Our consultants utilize a proven framework for what we term as a data scan – a process that identifies data silos, inconsistencies and gaps in your Engineering & Maintenance, Inventory data using our database and experience in dealing with multiple Fleet types (Narrow & Wide Bodies), Engines and MRO systems. This is the first step towards building the single source of truth database.
The result of a data scan is a recommendation report that elaborates on data gaps, process challenges and opportunities to unearth valuable insights from your data sets after enrichment in line with your goals. We can then assist you with the process of implementing the recommendations.
