Search archive          Sign up for our Newsletters          Aviation Jobs
Latest Aviation News  |  Industry & Technology  |  Air Transport  |  MRO & Support  |  Aircraft Interiors  |  Editorials  |  Events Calendar  |  About UsFR
 
Aviation News Thales applies predictive maintenance to military electronics

Thales applies predictive maintenance to military electronics

Léo Barnier
18 OCT 2018 | 1215 words
Thales applies predictive maintenance to military electronics
© Thales / E. Raz / Regards
Thales has been setting itself a double challenge for the last two years in favour of the French air force: to implement predictive maintenance in the military sector - which is much less advanced than the civilian sector in this area - and apply it to electronic equipment - an area which is said to be incompatible with this type of solution. While the system still needs several months to mature, with the introduction of new parameters to improve its reliability in particular, the first results are encouraging. The French electronics specialist is not slow to point out that it is a trail-blazer.

A first model was presented at the ADS Show, which was held at BA 106 in Bordeaux-Mérignac on 26th and 27th September to illustrate the progress in the work carried out by Thales's "data scientists" and technical experts. Since the start of the year, this model has been used to supervise the predictions made by the platform to check whether they came to fruition or not, and so test the performance of the algorithms, which are still at the experimental stage. The equipment provider is currently claiming an 80% success rate for the predictions.

The Hums contribution

The ambition of course is to achieve 100 %. While the objective is not possible in itself, the Thales teams should be getting very close to it thanks to the use of health and usage monitoring sensors (Hums) integrated into the electronic components, like what is already being done with mechanical equipment, but with a much higher level of accuracy. Although the sensors are already in place the data generated is not yet recorded and stored.

This will be the aim over the coming months for Thales, which is currently carrying out an experiment with the DGA with three cards on the French air force Rafale's RBE2 AESA radars. These collect data such as pressure, temperature, vibration and other information on the active antenna. Around 0.1 GB of information is collected with each flight. This Hums data is added to the data which is already being exploited.

For the moment, Thales uses around sixty parameters from three types of data. First of all, the data provided by the air force from the Harpagon system, which enables Rafale flight parameters to be collected and analysed after a mission and integrated equipment maintenance information stored by the plane.

Secondly, the ATAMS (Aircraft technical and airworthiness management system) data from the air force's technical and logistical information system, which logs the events which occur (as does Harpagon) but also all of the resulting maintenance actions.

Third of all, the data collected internally from Thales's test and production benches. While this does not fully show the reality of the operations, this last category enables a considerable volume of data to be available - around 12 000 GB - in comparison to the other two categories, which account for 1 GB and 1.4 GB respectively collected during 145 000 flights.


The working platform is used to arbitrate between the predicted volume of removals and the alert precision. © Thales

Probabilities and alert thresholds

For now, the algorithms are configured to establish the probability of failures every five flights for each item of equipment monitored within the fleet. The ultimate objective is to move towards longer intervals, every 15 or 20 flights of even every six months. The operator may then define an alert threshold which is more or less strict to remove an item of equipment. The model enables the operator to choose between 50%, 70% and 90% probability of a failure during the next five flights.

The first choice, 50%, guarantees a high level of reliability and above all enables operational availability to be preserved with the equipment which is the most able to carry out a mission. However, it does run the risk of excessive maintenance, compromising the optimum use of the potential of certain components, and so additional costs.

Conversely the third choice, 90%, will enable the component's maximum potential to be achieved and optimise the logistical chain with minimum stock levels and a just in time system. The other side of the coin is an increased risk of failure which may potentially compromise a mission. The choice of threshold will therefore depend on the context in which the aircraft are used, between OPEX or a training mission in mainland France, for example.

The objectives for using predictive maintenance in electronics are therefore the same as for mechanical equipment. The aim is to improve aircraft operational availability, increase mission success and optimise the supply chain.

Increasing accuracy and maturity

The increased accuracy in predictions - in particular with Hums, but also other elements such as maintenance reports via automatic natural language processing (NLP) - is vital to be able to exploit the platform's capacities to the maximum and so provide operational benefits. The research teams are therefore working on a certain number of performance indicators (KPI) to improve their algorithms to be able to predict all removals with the maximum level of accuracy. In-house, Thales feels that predicting 90% of removals with 90% accuracy would already be a very good performance.

The system should start to be gradually rolled out from Thales over the course of the next year. If the experiment with the DGA is a success, the industrial company will try to persuade its contacts to extend it to the whole Rafale fleet to raise the system's maturity level (which requires a certain volume of data and so equipment covered), but also to integrate it with a minimum amount of constraints for military mechanics (see box). A first operational version could then be proposed to the air force at the end of 2019.

But the platform's development shouldn't stop there. Thales wants to achieve a high enough level of accuracy for the system to be able to overcome alert thresholds and establish a "binary" response according to the types of mission (with predefined profiles): the equipment is fit for this mission or not. By 2020, the platform could event integrate logistics chain constraints and human and material resource availabilities into its removal proposals.


The platform is used to monitor equipment and recommend its removal. © Thales

Flying hour support offer

In commercial terms, Thales's aim is to manage to include this service in support contracts. The electronics specialised would then be able to offer a customer - France only first of all, even if the door is not closed to export customers - a fixed-rate commitment to the usage potential of their equipment, or even cover any additional cost in the event of an early breakdown.

In addition to the RBE2 AESA, where work is more advanced at the moment, Thales is working on two other items of Rafale equipment, the heads-up display (HUD) and the BACHD, the electronic warfare equipment integrated into the Spectra suite. And while the Dassault Aviation combat plane is currently at the centre of the experiment, the equipment provider is planning to extend its predictive maintenance platform to other aircraft such as the Tigre and the ATL2 or even to ships, in particular multi-mission frigates (FREMM).
Léo Barnier
Specialized journalist
Industry & Technology, Equipments, MRO


 
They made this section possible
Top stories
22 MAY 2020
Focus on the Airbus solution to convert passenger aircraft into auxiliary freighters Focus on the Airbus solution to convert passenger aircraft into auxiliary freighters
Editor's note : this article was first published on the french version of Le Journal de l'Aviation on April 30 Despite its own difficulties, Airbus continues ... Continue Reading
22 MAY 2020
AKKA Technologies partners with Avianor to offer an EASA approved temporary conversion solution for wide-body aircraft AKKA Technologies partners with Avianor to offer an EASA approved temporary conversion solution for wide-body aircraft
AKKA Technologies and Avianor have signed a framework agreement aimed at offering a temporary Passenger-to-Freighter conversion solution for airlines that require EASA (European Aviation ... Continue Reading
20 MAY 2020
MRO: Asia's largest maintenance hangar enters service at Beijing-Daxing airport MRO: Asia's largest maintenance hangar enters service at Beijing-Daxing airport
A few months after launching its line maintenance activity at Beijing-Daxing International Airport (PKX), GAMECO (Guangzhou Aircraft Maintenance Engineering Co. Ltd) is now just starting ... Continue Reading
13 HOURS AGO
LATAM files for bankruptcy in the US
25 MAY 2020
Berlin announces temporary part-nationalisation for Lufthansa
23 MAY 2020
A Pakistan International Airlines A320 crashes near Karachi (updated 23/05)
22 MAY 2020
Ryanair's Laudamotion says closing Vienna base, 300 jobs lost
22 MAY 2020
Focus on the Airbus solution to convert passenger aircraft into auxiliary freighters
Top stories
Focus on the Airbus solution to convert passenger aircraft into auxiliary freighters
AKKA Technologies partners with Avianor to offer an EASA approved temporary conversion solution for wide-body aircraft
MRO: Asia's largest maintenance hangar enters service at Beijing-Daxing airport
How Air France manages its grounded fleet before recovery time
Rolls-Royce aircraft engines at idle
 
Latest News     Industry & Technology     Air Transport     MRO & Support     Aircraft Interiors     Editorials
© 2020 Le Journal de l'Aviation - All rights reserved