Leveraging predictive analytics towards in-flight fuel savings
Author: Delphine Guérin, Head of Flight Efficiency Department, Air France; Kim Heap, Project Manager – Flight Efficiency, Air France; Elisabeth Monteiller, Europe Sales Manager, Safety LineSubscribe
Air France has implemented a data-based fuel efficiency system to manage fuel and reduce emissions.
Before we consider why and how OptiFlight was deployed at Air France, and before we look at the details, we’ll give you a very brief introduction to Air France’s environmental commitment and the flight efficiency program (figure 1).
Reducing Air France’s CO2 emissions goes hand-in-hand with reducing fuel use. Therefore, one of the cornerstones of the airline’s environmental policy is to reduce the use of fuel. Air France is continuously investing in new equipment; a modern fleet not only provides greater comfort to passengers, but also helps to achieve significant fuel saving and to meet the airline’s sustainable development commitments by reducing CO2 emissions and noise levels. One area where Air France is making a difference is in the promotion and use of sustainable aviation fuels; an important factor in emissions reduction. Both Air France and sister airline KLM have policies in place that encourage and support the development and use of alternative fuels. However, currently there are not enough supplies available to allow a full switch to sustainable alternative fuels; therefore, carbon offsetting is a further way to offset the airline’s environmental footprint. Last, but not least, improving flight operations and implementing weight reduction measures on board, optimizing flight routes and using up-to-date weather information all contribute to the reduction of fuel used.
At Air France, there have been fuel plans in operation since 2009. The first actions to improve fuel consumption were implemented between 2012 and 2014. It was decided to generate a fuel plan during 2016-17 (figure 2) which resulted, over two years, in a reduction of CO2 emissions by 60,000 tonnes per year. Subject to the strict rules of flight safety, every possible fuel saving measure is identified and implemented at Air France. All Air France departments are involved in the fuel plan, and the aim is to reduce fuel consumption by improving the operations processes, by making partnerships and innovating in the supply chain and by mobilizing staff and the industry.
Now, we’ll share with readers the reasons why OptiFlight was deployed at Air France and how that was done. To ensure better weather awareness for flights, Air France together with SITA FOR AIRCRAFT has developed eWAS. The development started in 2017 and has since been upgraded (figure 3).
CONVEYING FLIGHT-SPECIFIC BEST PRACTICES
The opportunity that came about in 2019 was to integrate OptiDirect advice with eWAS and the same for the other OptiFlight modules. What happens with OptiFlight takes fuel efficiency to the next level (figure 4).
Most airlines are familiar with Excel sheet fuel dashboards that offer descriptive analytics: but, importantly, compared to those solutions that only analyze best practices, OptiFlight takes fuel efficiency two steps further on by using predictive analytics with machine learning performance models that allow users to predict fuel consumption in different scenarios and to prescribe recommendations directly to pilots to enable them to select the best scenario for each flight. That is achieved by bringing data science into the cockpit; leveraging historical flight data towards developing flight specific pilot recommendations. Historical data is used to build machine learning performance models for each tail number and then that is combined with the OFP (Operational Flight Plan) data and the 4D weather forecast of the day. Next, a ‘what-if’ engine is run considering tens of thousands of different scenarios to identify the best scenario and send that, by the simplest way, to the pilot. It can be integrated into OptiBriefing or also be directly available into the eWAS Pilot App.
What happens, more specifically, with OptiDirect (figure 6) is that the historical flight data of the user company is matched with the global waypoints database. That way, by using machine learning, the system is able to identify which waypoints the actual trajectory went through.
That is done because the actual trajectories, shown as white on the chart, are quite different from the OFP data that is shown in red. For that reason, OptiDirect uses machine learning to identify each waypoint of the actual trajectory and then connecting the dots, waypoint by waypoint, to re-create those trajectories and then build the historical flight track database of the company.
The next step is to deliver shortcut opportunities based on those historical tracks flown (figure 7).
This allows pilots to do what they have always done when requesting directs from the ATC to save time and fuel, but to do that better. They have not previously had the right information to be able to determine which direct they should take or how much saving they will achieve through that choice. They didn’t really have those answers and that’s why airlines use OptiDirect to implement shortcuts based on historical tracks flown with an indication of fuel and time savings.
In the figure is an example from OptiDirect between the two Waypoints RANUX and ARCKY that would save 71kg of fuel and 2 minutes flight time; the information also indicates the number of times that the direct has been flown before, nine times in this case, and with this information, pilots can be confident that if they request this direct from the ATC, they might get it again.
OptiDirect received a new patent in June 2020 and that solution has already been implemented at Transavia, Air France and at AeroLogic, representing more than 300 aircraft and up to 900 flights a day
In order to get ready to implement and to test OptiDirect, Air France started a two months trial with a group of fifty key users, pilots, who continuously provided feedback, positive or negative, related either to OptiDirect or to the integration with eWAS. A shortcuts database was built for the trial, based on six months’ historical ADS-B data covering the Air France network. The trial involved pilots from the whole Air France fleet except the A380 aircraft.
During the two-month trial, the OptiDirect module was tailored to deliver the most relevant advices. That was the reason for excluding any direct proposal below flight level 120. At this stage, there had to be limitations in order to display the five best direct advices, otherwise the flight path was too overloaded with advices. Associated with it, there was a minimum saving of 20kg of fuel per direct required. Similarly, the maximum number of directs per waypoint was limited to one. Also, there was a limitation on the length of direct. And, finally, at least six occurrences within the past month must be observed in order to be displayed.
During this trial, due to ADS-B data precision, there was some OptiDirect advice that took flights too close to no-fly zone areas (figure 8).
Air France mentioned that to Safety Line and they worked together towards integrating the no-fly zones specific to Air France as the implementation evolved over time. An example, would be when OptiDirect advice was to overfly the Ukraine no-fly zone area: after pilot feedback provided the information, Safety Line quickly adapted the OptiDirect engine.
Another challenge faced for medium-haul flights, was that OptiDirects advices cover multiple areas that require an ATC co-ordination. As an example, flights out of Paris to destinations in Italy. On the other hand, a positive thing, illustrated in figure 9 is a flight out of an area where Air France doesn’t often request a new direct. This direct was requested by the crew and they got it.
Thanks to the trial campaign, Air France was able to build a strong business case for OptiDirect. Over the two months of the trial, more than 146 directs were proposed for each aircraft every month, with an average saving per direct proposed of 65kg which led to a breakeven being reached if only 1.9 percent of directs are granted.
Since OptiDirect is statistics driven, the first engine was based on ADS-B data which, depending on the context, is more or less accurate. To advance it, Safety Line and Air France are working on adding QAR data with much better data quality in terms of precision and data sampling.
Air France expects its partnership with Safety Line and the OptiFlight modules to further improve fuel use rates.
To support Engine Out Taxi Out procedures, Air France embedded the advices in eWAS, attached to the departing airport and, if applicable, the arrival airport (figure 10).
The OptiDirect module is available with eWAS, an operation application widely used by Air France pilots. Using the OptiDirect module is intuitive (figure 11). Once the flight has been selected in the flight planning manager, a green leaf icon along the route, identifies where short-cut information is available.
By clicking on the icon, a window opens and gives short-cut details, waypoints, fuel, flight-time savings and the number of times this short-cut has been flown. A green dotted line also displays the short-cut on the map. eWAS integration allows access to information without having to open another application or document. Integration will be the key for the coming years with a large number of products and onboard innovations. The icon, at the arrival airport will give information on applicable operating procedures and associated fuel savings.
CO-INNOVATING IN FUEL EFFICIENCY
In spite of these difficult times, both Air France and Safety Line are committed to reducing the environmental footprint of aviation and that’s why they are co-innovating to improve fuel efficiency. Air France is an innovation partner with Safety Line which means that the two are exploring together further innovation opportunities such as OptiClimb which is an App for flight optimization in the climb, OptiLevel to optimize flight levels during the cruise and OptiDescent which helps pilots anticipate the most likely approach plan. Also, collaborating with eWAS pilots, a Green Operating Practices (GOP) module has been developed (figure 12).
GOP module is a new eco-flying module available in eWAS where the pilot can find all fuel efficiency best practices tailer-made for each flight that covers all flight phases.
SITA acquires Safety Line and takes the lead in deploying predictive data analytics to reduce CO2 emissions and improve fuel savings.
SITA, the global IT provider for the air transport industry announced, in July 2021, the acquisition of Safety Line S.A.S., the Paris based start-up specializing in digital solutions for aviation safety and efficiency. This acquisition will strengthen SITA’s Digital Day of Operations portfolio, helping airlines drive more efficiencies and fuel savings around the aircraft while taking immediate and sustainable steps to reduce their carbon footprint.
With air transport accounting for about 3% of the worldwide carbon emissions, there is growing pressure on airlines to reduce their overall emissions. At the same time the COVID-19 pandemic requires airlines to make their aircraft operations leaner, in particular reducing costly fuel burn. Safety Line has successfully applied predictive analytics to deliver significant improvements to aircraft operations, strongly complementing SITA’s existing portfolio. With this acquisition, SITA will accelerate the development of sustainable solutions that can be integrated with its existing suite of airline and airport solutions.
Reducing aircraft CO2 emissions in-flight and at airports
The acquisition comes as a logical follow-up to the commercial partnership signed between the two companies in September 2020, focused on helping airlines reduce their CO2 emissions. OptiFlight is the only predictive in-flight fuel efficiency solution leveraging machine learning performance models for each aircraft to optimize all flight phases including climb-out, cruise and descent. OptiFlight is already fully integrated with existing applications in SITA’s Digital Day of Operations portfolio, and current joint customers include Air France, Transavia Airlines, Aerologic and Condor.
On the ground, Safety Line’s AirsideWatch will allow SITA to expand its airports offering to airside operations, using surface movement radar data to better understand, analyze and optimize the ground traffic of aircraft between gates and runways with an aim to reduce unnecessary emissions.
Using data to improve safety and risk management
With safety being a key pillar in aviation alongside efficiency, the acquisition also covers SafetyCube, a solution that enables aviation stakeholders to pro-actively manage their safety and compliance. Key customers include large airports such as Paris-CDG and Paris-Orly or OEMs such as Airbus Helicopters, as well as airlines and ATM providers, all of which also compose SITA’s main customer base. SITA will, therefore, be able to expand its offering to include safety-enhancing services to existing customers.
Delphine Guerin is the Head of Air France Flight Support and Efficiency Department, a position she has held since 2017. A graduate from a French Aeronautical Engineering School, Delphine Guerin joined Air France in 2004 and has had professional experiences in Ground Operations, Dispatch, OCC and Flight Operations. She has taken part in many key Air France projects such as the opening of S3 and S4 terminals at Roissy Charles de Gaulle Airport or the integration of LIDO flight planning.
Kim Heap is a project manager specialized in Flight Efficiency, a position he has held since 2019. A graduate from French engineering school (ECE Paris), Kim Heap joined Air France in 2013 and has had experiences in Freight and Flight Operations.
Elisabeth Monteiller is Safety Line regional sales manager in charge of EMEA region, a position she has held since 2020. After graduating from ESCP Europe French business school, she started her career at Textron Aviation, as part of Textron’s Sales Development program. She had professional experiences in Business aviation aircraft sales and aftermarket sales. Before joining Safety Line in March 2020, she was aftermarket sales associate for Cessna Service Center based at Le Bourget, Paris, France.
SITA FOR AIRCRAFT
SITA is the IT provider for the air transport industry and is 100% owned by the industry, delivering solutions for airlines, airports, aircraft and governments. SITA FOR AIRCRAFT is powering a digital shift to make air travel more connected, seamless, efficient, safe and sustainable. Its communications network connects every corner of the globe and handles vast volumes of data every second.