Future of MRO
When Big Data is used for maintenance
To date, maintenance operations are “driven” by statistical approaches (preventive maintenance) and, more recently, by predictive approaches based on Big Data and AI. These different methods aim to anticipate problems to minimize aircraft downtime. However, these approaches all have their limits, especially when the human factor intrinsic to maintenance must be considered.
The traditional maintenance approach
The traditional approach through statistical analysis consists in calculating the probability of “mechanical” failure (FMECA type) to anticipate the replacement of a part before it is defective.
To do this, a set of procedures (tasks) are listed in a maintenance plan (MPD) to be executed according to a precise schedule. The MPD is a document that includes scheduled maintenance as well as the entire aircraft maintenance policy. This mandatory and regularly updated maintenance plan lists all the tasks to be performed, expressed in flight hours and/or in number of cycles and schedules.
Today, platforms are made available to airlines by manufacturers or suppliers (Airbus Skywise, Lufthansa Technik AVIATAR, etc.). The purpose of these platforms is to have as much information as possible for a given type of aircraft, to obtain the widest possible coverage of incidents/events.
The limits of the planned approach
Based on planned maintenance approaches, for example, a C-check for an aircraft includes about 1000 tasks and 5000 planned sub-tasks. However, during the execution of this C-check, between 500 and 1000 additional unplanned tasks will have to be executed. This is the limit of this planned maintenance approach, which can over- or underestimate certain tasks.
Moreover, the disadvantage of this type of approach is that it does not consider the particularities of each airline, nor the operating conditions. The impact of pollution on the ageing of a fleet will be different in the Far North than in the Middle East or China, for example.
The predictive approach, a new approach still in its infancy
Predictive maintenance consists of considering a large amount of information generated by the aircraft via real-time message mechanisms (ACARS type) in order to detect a set of events that could be the cause of a failure (Root Cause).
The objective is therefore to recover real and dynamic external data from sensors to enrich the field feedback and allow a more detailed analysis. However, the predictive approach also has its limits, as it is highly dependent on the quality of the sensors available on the aircraft (new A350-B787 programs versus older A320-B737 programs) to collect the necessary data.
The solution developed by Fingermind now makes it possible to compensate for the quality of the sensors to facilitate and enrich the recovery of “physical” data on the ground.
Fingermind’s MRO Suite is thus able to collect and centralize very specific data such as temperature, luminosity, a large set of contextual information… Moreover, this collection is possible both online and offline.
MRO Suite will then extract a set of “nominal” and “learning” data to generate an AI model and transfer the model to the mechanic’s device.
The human factor, the great forgotten factor in MRO
Both the scheduled and the predictive approaches unfortunately ignore human factors as well as the use of tools in the field to access technical documentation in each situation.
Thus, these approaches to maintenance do not consider the way in which maintenance tasks are carried out, the capacity and expertise of the mechanic, the difficulties he may encounter, the complexity of the tasks, the context of use of the tools, the climatic conditions, the working hours (night, day…), etc.
However, the way of executes the various tasks necessarily has an impact on the maintenance of the aircraft and especially on the time spent by the mechanic.
It is therefore essential to understand the working environment and the behavior of the mechanic in order to provide additional information and to allow better decision making to optimize the maintenance procedures by the airline company.
A collaborative working group to go further
Today, Fingermind wants to respond to this challenge by bringing together airlines, mechanics, engineers and all MRO stakeholders within a working group to co-develop solutions capable of better collecting and analyzing data and facilitating decision making.
The aim is obviously not to incriminate this or that mechanic, or to point out a possible lack of training of the teams. The subject must be treated in its entirety from the point of view of improving existing processes, taking into account all the interactions in the field.
Various studies have already listed several human factors as possible causes of malfunctions during maintenance operations:
- Lack of communication within the teams
- The routine nature of certain operations
- Lack of competence despite the different mandatory certifications
- Lack of concentration during the execution of a task
- The difficulty of working in a team
- Fatigue and schedules
- The pressure of responsibilities
The objective of the working group is to help the mechanic to better use the documentation to perform the different maintenance tasks within the time limit:
- While respecting the safety rules for the mechanics
- While guaranteeing operational safety for the aircraft
- With a better control of the execution time (planification)
- With a concern for traceability and quality in the collection and restitution of data
More globally with the aim of improving maintenance processes.
Fingermind has therefore launched a LinkedIn group called “Future of MRO” to bring together stakeholders and give them a voice in their respective expectations and challenges. This joint reflection could ideally lead to a pilot solution or a beta test to propose solutions that meet the needs of MRO players in a concrete way.
If you too would like to participate in this group to make progress together in MRO processes, come and join us here!
The Fingermind teams and I will share our various thoughts, developments and advances within this group to discuss their usefulness and feasibility.