Artificial intelligence, data science and predictive maintenance" are already technologies used by various players in the sector. As the civil aviation authority, we also use these technologies, particularly to optimise our surveillance plans," explains Hugues Carrière, Director of Methods and Development at OSAC. Indeed, thanks to the power of data analysis, the focus is now on prioritisation and prediction.
Algorithms
in the field
By cross-referencing reports from the Civil Aviation Safety Investigation and Analysis Bureau (BEA), safety notifications and the results of audits by aviation organisations or inspections, OSAC can prioritise its inspections where the risks are considered to be highest. "We identify weak signals. The aim is to anticipate rather than react," adds Hugues Carrière. SixFoisSept, a subsidiary of the Apave Group specialising in smart data and AI, supports this strategy. “Each aircraft has a non-compliance risk score, broken down by area: maintenance, documentation, cyber security, etc.,” explains its founder, Erwan Prud'homme.
Fed by multiple data sources (audits, incidents, organisational and financial informa- tion), this score makes it possible to objectify decisions, adjust the frequency or depth of controls, and above all, better allocate scarce resources. All without a black box: "All our algorithms are explainable and approved by business experts," he insists.
Supporting
the revolution
Applications are diversifying: automated visual inspection, AI-analysed engines, chatbots facilitating the search for technical documents, etc. These are powerful tools, but they need to be regulated. "The EASA is preparing regu- lations on these technologies. The chal- lenge for us is not to slow down innovation, but to make it safer," says Hugues Car- rière. "Humans remain in control. But with the right tools, they can see further and fly safer," concludes Erwan Prud'homme.
Sémaphore Risk Monitoring©:
a revolution in aeronautical surveillance
How do you effectively inspect 14,000 aircraft every year? With Sémaphore Risk Monitoring© , Apave offers an innovative approach based on detailed risk analysis. This risk calculation engine optimises inspection plans guarantee even greater and modulates surveillance to guarantee safety even further.
Managing the monitoring of an aircraft fleet requires a rigorous strategy. To meet this challenge, the Apave Group has developed Sémaphore Risk Monitoring©, a data science tool deployed in production at OSAC. The principle behind it is to calculate the risk of non-compliance for each aircraft, for each registration, based on past audits, non-compliances identified and their resolution times. "Sémaphore doesn't just identify the riskiest aircraft: it also takes into account business constraints to ensure comprehensive coverage of aircraft types and maintenance workshops," explains Erwan Prud'homme. The result is an ‘ideal list’ of aircraft to be inspected.
Since 2019, Sémaphore Risk Monitoring© has also made it possible to adjust the frequency and scope of inspections. An efficient maintenance workshop can now space out its checks from two to four years, while a struggling organisation will be monitored more closely. Each area (human resources, tools, archiving, etc.) is assessed inde- pendently for the most accurate management possible. "What makes this tool unique is that it is based on an explainable calcu- lation engine, without the use of opaque artificial intelligence algorithms. Successfully audited by EASA in 2024, Sémaphore Risk Monitoring© is also used in the nuclear industry and was recognised as one of the three best innovations in operational excel- lence at WNE 2023."
A synergy serving flight safety
Discover our
news