Royal Navy looks to algorithms and analytics to detect anomalies in helicopters

Ministry signs contract with specialist firm to provide an application for engineers to probe data

The Wildcat HMA2 will be among the craft analysed by the anomaly detection tool    Credit: Crown Copyright/Open Government Licence

The Royal Navy is progressing its use of algorithms and analytics to spot potential engineering issues with its fleet of Wildcat helicopters.

Newly published commercial documents reveal that the Ministry of Defence has signed a one-year deal with a specialist firm that will support the development of an “anomaly detection analysis and reporting tool”.

The contract reveals that, as part of the Navy’s Programme Nelson tech transformation scheme, an algorithm was developed in 2019 to detect anomalies in the functioning of helicopter components such as “gearboxes, drive train, engines and rotor systems”. 

This algorithm is fed by information from a monitoring system that collects data from on-craft vibration sensors. These “raw vibration signals generate condition indicators which are indicative of the health of” helicopter systems – including “information relating to component wear and damage”. 

The monitoring platform identifies instances where vibrations exceed a stipulated threshold, potentially indicating attrition or damage. These can then by further investigated by engineers.

Wildcats – which in 2015 began to replace the Navy’s previous generation Lyn attack helicopters – are equipped to provide engineers with 400 individual condition indicators (CI), which the contract says “makes manual analysis and aircraft monitoring highly challenging, especially on a fleetwide basis”.


Related content


“While existing automated alerts can be highly effective in detecting certain fault conditions, this approach can be less effective at alerting for other subtle, rapidly progressing faults. This is where the complementary approach of anomaly detection is beneficial,” it adds. “Through the use of an anomaly detection algorithm, which compares the behaviour of each individual CI for the component, rapidly developing fault conditions may be identified before component failure and before an exceedance alert is triggered. This project seeks to procure an automated CI data analysis capability using anomaly detection, which will help identify early onset of component damage. This would enable informed monitoring and fault-management using other established methods such as wear debris analysis and lubricating oil analysis.”

The algorithm developed by the Navy in 2019 “was proven to have success”, the contract says. The intention now is to take this tool and “develop and incorporate it into a software application that is useable by frontline aircraft engineers”.

This work will be delivered by Tonic Analytics, a software firm specialised in big data and predictive analytics. The company signed a 12-month contract with the MoD on 1 October; the deal will be worth £45,000 to the Southampton-based company.

The firm will be tasked with taking the Python-based code of the anomaly detector created by Programme Nelson and refining and improving its functionality. 

“The coding should then be incorporated into a Windows or web application-compatible solution with an appropriate graphical user interface,” the contract says.

Throughout the process, work should be informed by input from the 1,710 Naval Air Squadron engineers that will use the software, the document adds.

The ultimate goal is that the automated means of flagging potential anomalies “will aid and prioritise deeper analysis” by both monitoring software and engineering personnel. 

The application should initially work with HMA2 and AH2 models, “with future capability for additional airframe types to be incorporated” in due course.

The MoD will retain all intellectual property rights for the software. 

 

Sam Trendall

Learn More →

One thought on “Royal Navy looks to algorithms and analytics to detect anomalies in helicopters

  1. 홍대출장섹스마사지 October 9, 2024 at 9:29 pm

    https://m.expert.naver.com/mobile/expert/product/detail?storeId=100032751

Leave a Reply

Your email address will not be published. Required fields are marked *

Processing...
Thank you! Your subscription has been confirmed. You'll hear from us soon.
Newsletter Signup
Receive the top stories from the UK’s leading public sector digital and data publication direct to your inbox every lunchtime.
ErrorHere