Crowdsourced data helping to keep aircraft flying
Aircraft manufacturers and suppliers are starting to use crowdsourced data to detect wear, schedule preventative maintenance and avoid costly failures. Steve Nichols reports.
The current thinking is, that by pooling sensor data across an airline, or even across an aircraft type, problems can be detected before they become a major issue.
The 2017 Connected Aircraft eEnablement Conference, held recently at Heathrow in the UK, heard how engine manufacturers now view the importance of gathering masses of data.
For example, Skywise from Airbus, which was launched at the Paris Air Show in June, brings together aviation data from multiple sources across the industry into one secure cloud-based platform.
These airline sources include work orders, spares, components aircraft/fleet configuration, on-board sensor data and flight schedules.
Jaime Baringo-Ezquerra, head of digital business development, Airbus, said: “What could we achieve if data from every Airbus aircraft could be gathered together? We have already build this capability within Airbus and within a few weeks we were able to see double-digit improvements in productivity.”
Airbus digital transformation leader, Eric Peyrucain, added: “By benchmarking your data against other operators, airlines can see how they are performing, or they can use other companies’ data collaboratively to predict failures.”
Airbus says Skywise includes 24,000 parameters with 100% availability of the data.
Boeing heads for the clouds
Boeing is also in partnership with Microsoft to adopt a “cloud computing” approach to capturing aircraft data.
The collaboration aims to improve commercial aviation by enhancing factors like predictive aircraft maintenance, fuel optimisation, and airline systems.
Alex Montgomery, cloud application lead, Microsoft, said: “Boeing’s partnership with Microsoft will accelerate innovation in areas such as predictive maintenance and flight optimisation, allowing airlines to drive down costs and improve efficiency.
“We need to think about supporting predictive failures, based on historical data. Increasingly, we are seeing machine learning and analytics being moved to the aircraft, to be closer to the problem.
“Now is the right time. With the unlimited capacity of the cloud and powerful analytical algorithms, airlines can now exploit the data they are capturing.”
Monitoring engine health
Engine manufacturers Rolls-Royce and GE are also looking at how big data can monitor engine performance and reliability.
Mike Chester, solution manager, operator solutions, digital, Rolls-Royce, spoke about how his company is using eEnablement for engine health monitoring.
It can gather thousands of data points from each flight and compare these with known norms to look for data points that are out of the ordinary and suggestive of a problem or impending failure.
Rolls-Royce has developed a predictive maintenance app that studies a number of parameters from the fuel, avionics, and hydraulic systems to look for potential problems. The app was developed and delivered in just 12 weeks.
“The complexity of aircraft data means we can’t afford to get it wrong,” Chester said.
“We have to be as accurate as we can. At Rolls-Royce we still need an engineer with dirt under his or her nails, who knows what a 5/16th socket is. But we also need analysts and scientists, who can interpret complex data and use machine-led algorithms to look for problems.”
Steve McFeely, chief engineer connectivity, Boeing, said: “It is predicted that more than 50% of organisations globally will use advanced analytics and proprietary algorithms to cause disruption of entire industries by 2018.
“For example, the Boeing 787 had an 80x increase in the number of data parameters over a 737. The 777X will show a 100x increase.
“The goal is to ensure that every aircraft that leaves our factory is connected, but the key is to ensure that they are cyber protected.”
John Nelson, operations manager at General Electric, said that getting real-time information off an aircraft was possible but you needed a “fat connectivity pipe” to the ground.
“The aircraft communications addressing and reporting system (ACARS) is still going to continue be used for some time,” he added.
He said using data from the aircraft once it has landed was still the way to go, at least for the time being. He said GE had been working with a major undisclosed Middle East customer that operated its aircraft in harsh environments.
“Better predictability over future problems has helped bring a 43% decrease in disruptions and 12 additional days of aircraft utilisation,” he said. “This reduces the maintenance burden considerably.”