Aircraft generate massive amounts of data and advances in processing have allowed the development of innovative solutions that can help to save fuel. One hindrance to these is crowded airspace, which mean altitude changes for more favourable winds, short cuts by direct routing, or continuous descent approaches are not always possible but it still possible to make other aspects of a flight more efficient. In addition, burning fuel to carry fuel can be reduced by more accurate planning.
François Chazelle, chief commercial officer, says Safety Line is celebrating its tenth anniversary this year. It started as a funded research project at the Telecom Paris Tech Incubator in 2011, with the company being founded in the same year. Its first product was SafetyCube, an integrated compliance, safety and risk management solution,
Analysis of some 30 parameters of flight data stored by the Quick Access Recorder led to the realisation that it very precisely captured individual aircraft performance in different configurations. An in-house data science research lab was set up to develop predictive and prescriptive solutions, using Machine Learning performance models for each aircraft, that could optimise specific flight phases and provide recommendations to pilots. There is an OEM aircraft model, he notes, that is used by the FMS but it is not as accurate as real life data, which can also detect changes in aircraft performance over time.
The first area of interest was the climb-out phase, which has the highest fuel burn. Often, climb-out starts with a first speed of 250kts to 100,000ft but this, he says, is a response to regulations that say a maximum of 250kts to a minimum of 100,000ft. The result was OptiClimb, launched in 2018, a more nuanced approach with the potential to produce significant savings. By using wind and temperature inputs every 1000ft, which are predictable up to 12 hours ahead, customized climb speeds and altitudes are sent to pilots to be included in their briefing package. These are slower horizontal speeds which for a given thrust result in higher vertical speeds and a higher climb angle, getting the aircraft to altitude sooner, where it consumes less fuel.
Transavia France was the first airline to implement OptiClimb. The first experimental flights started in Summer 2015, followed by an initial test phase with 10% of the pilots the following year. In the second half of 2017, it was tested by sister company Transavia Netherlands. A contract was signed with both companies in December that year. In 2019, Transavia France had 18,067 OptiClimb flights, with an average saving of 85kg of fuel per flight, giving a reduction in CO2 emissions of 4,837 tons.
Mexican ultra low-cost carrier VivaAerobus started testing OptiClimb in April 2020. This followed a full scale trial of almost 4000 flights before the pandemic eventually reached Mexico, with application rates as high as 83%, providing plenty of useful data. As a result, the airline should be able to save in average of 70kg of fuel for each climb, which could represent a fleetwide carbon footprint reduction of at least 14,000 tons of CO2 per year.
Other OptiClimb customers include Air Austral, Air Asia, Sky, TAP Express and TUI.
OptiClimb is now just one of a suite of programs that form OptiFlight. The others are:
• OptiSpeed, which shows pilots the fuel and time impact of Mach variations to enable on-time arrival with the best fuel/time ratio.
• OptiDirect, which recommends shortcuts that pilots can request from ATC based on historical tracks flown, with an indication of fuel and time savings taking into account the wind and temperatures forecasts for the flight.
• OptiLevel, which advises pilots on the best initial flight level and potential cruise level changes taking winds into account.
These three products are packaged as OptiCruise.
Finally, there is:
• OptiDescent, which helps pilots better anticipate on Distance to Go based on Machine Learning of historical approach patterns, including landing direction and the time of day.
OptiDirect trials were carried by Transavia, starting in June 2019. Almost 18 months later, pilots had taken 1895 shortcuts, with average savings of 37kg of fuel and 55 seconds of time saved on average per shortcut. That represents 84tons of fuel and 35 hours of flight time saved in the test, with CO2 emissions reduced by 264 tons.
Other OptiDirect customers include Air France, Aerologic and Condor.
Following a partnership agreement in September 2020, OptiCruise was integrated in SITA’s widely used eWAS Pilot mobile application, which is part of SITA FOR AIRCRAFT’s ‘Digital Day of Operations’ portfolio. eWAS Pilot, used by 50,000 pilots of commercial airlines, business jets and cargo airlines, provides accurate 4D weather forecasts and real-time updates from various sources to warn about weather hazards such as thunderstorms, lightning, clear air turbulence, strong winds, icing and even volcanic ash.
First customer for the new package was AeroLogic, the joint venture between DHL and Lufthansa Cargo. It operates around 12,000 international flights a year with a fleet of 17 Boeing 777F freighters.
In July 2021, the link between the two companies was strengthened when SITA announced the acquisition of Safety Line.
This will bring Safety Line’s AirsideWatch into SITA’s portfolio, expanding its airports offering to airside operations
AirsideWatch uses surface movement radar data. This is usually used to monitor live ground traffic of aircraft and airside vehicles for safety purposes. However, by converting the data into searchable aircraft trajectories, it provides insight on a variety of criteria such as multiple points of passage, airline, aircraft type, date and time, type of trajectory phase, and visibility and lighting conditions.
Trajectories are broken down into specific phases, from parking and pushback to line up and take off, or from landing and runway exit to the gate, with the possibility to identify the time and distance covered. This allows for additional analytics such as taxi time and distance, time at gate, or time at de-icing bays. These extremely precise inputs can be incorporated into noise and emissions simulation models to help airports to reduce their environmental impact.
Finally, Safety Line has been involved in several two and development projects related to the reduction of CO2 emissions. As part of the European Commission’s Clean Sky2 research program, PERF-AI will apply Machine Learning techniques on flight data to accurately measure actual aircraft performance and provide real time optimisation of flight. It is joined by INRIA Lille, the French national research institute for digital sciences, and Thales as Topic Leader and had Lufthansa and Transavia France on the project advisory board.
An earlier project also involved INRIA, this time the COMMANDS research team from the Saclay–Île-de-France research center, whose main focus is studying dynamic optimization. This has seen the creation of an INRIA joint Innovation Lab called OptimiSation of Consumption for AiRplanes (OSCAR). The three-year project aimed to improve climb optimization.
LATAM DIVES INTO THE GREEN
LATAM Airlines is to upgrade over 200 of their A320 Family fleet by adding the Descent Profile Optimisation (DPO) function from Airbus to aircraft’s Flight Management System (FMS) performance database. All the equipment kits required for the installation of the DPO performance software will start to be delivered from the end of 2021 until early 2022.
The DPO function allows aircraft to descend from cruise altitude using only idle engine thrust, which reduces fuel consumption, bringing proportional CO2 and NOx reductions. It also maximises the time spent at efficient cruise levels by not starting the descent too early, which minimises the amount of time spent at the inefficient ‘level-off’ stage at the bottom of the descent, when the aircraft’s engines need higher thrust to maintain level flight in dense air prior to final approach.
LATAM Airlines will generate savings of more than 100 tons of fuel and more than over 300 tons of CO2 emissions per aircraft per year across their network, including constrained airports like Lima, Santiago and São Paulo. Across the fleet, this represents a reduction in fuel consumption of more than 20,000 tons and 60,000 tons of CO2 emissions.
TAKING CONTROL OF EMISSIONS
While saying that airspace efficiencies are sometimes compromised by ATC restrictions, it is only fair to look at the experience of one airspace navigation services provider. In this case, it is NATS in the UK, which has handled up to 2.4 million flights and 250 million passengers in a year.
Obviously, that was before COVID-19, which, ironically, made it easier to be more efficient with a reduced level of flights – holding patterns almost entirely vanished, as did vertical limitations, direct routings increased, and continuous climbs (direct to 100,000ft without levelling off) increased by 15% to 85% of all departures. For arrivals, the biggest fuel saver is Continuous Descent Approaches (CDA), a smooth descent with reduced power and no levelling off. However, this was unaffected by the pandemic. In fact, the 2019/20 average was 80% across the 22 UK airports covered by NATS, with London airfields operating around 90%.
The focus now is to maintain that efficiency as the industry recovers, although, in the case of CDAs, higher targets will only produce marginal gains as 100% is impossible due to factors such as go-arounds and non-standard conditions such as high winds.
NATS also gathers other data such as continuous climb, fuel burn and CO2 of the actual radar tracks and airline flight plan, as well as the amount of track extension over optimum. This is used in what it calls it 3Di metric (three-dimensional insight score), which has been running since 2012 as part of the contract with the UK Civil Aviation Authority (CAA). This measures the efficiency of every commercial aircraft flight, which, across the year, are averaged and compared to targets set by the CAA. While it is a barometer of performance, with financial penalties or bonuses, NATS says it is also a strong incentive to make every flight handled as efficient as possible, with subsequent reductions in emissions. The data is used by airports, airlines and ATC to identify differences and opportunities.
The U. S. company has taken a slightly different approach to emissions reduction with its Honeywell Forge Flight Efficiency program as it allows flight planners to reduce the amount of reserve fuel that needs to be carried, while meeting regulatory safety minima.
Traditionally, fuel loads are determined using flight plans, weather forecasts, navigation changes and aircraft performance data as well as historical fuel burn records on each sector. This is known as Statistical Contingency Fuel (SCF). However, airlines have always been very conservative in their fuel loading practices.
In 2015, environmental researchers studying a major US airline found 4.48% of the fuel consumed on an average flight was due to carrying unused fuel, with 1.04% consumed to carry fuel above what they called “a reasonable buffer.” That is important as, depending on the aircraft type and configuration, it takes 3-4kg of fuel per hour to carry each 100kg of load.
The company estimates that SCF quantities can be reduced from today’s customary 5-10% to the range of 3% or even less, saving hundreds of kilos of weight and dramatically reducing fuel consumption on a typical flight.
For a typical airline with a mixed fleet, the most frequent SCF exceedances for flights less than 500 miles and over 1,500 miles would be in the 3-3.3% range, with 4.5% for flights between 500-1500 Very long flights have a wider discrepancy (5.5%) deviation.
The variation is often greatest when flight times between a particular city pair may be affected by such things as headwinds and tailwinds. For example, there may be a smaller exceedance on an outbound flight between Chicago and Phoenix than on the return flight. The flight planner might allocate an additional 3% of fuel for each leg, whereas Honeywell Forge Flight Efficiency would probably recommend 1% for the outbound flight and 5% for the return. While that might mean refuelling at Phoenix, there are still overall reductions in costs, fuel consumption and CO2 emissions.
“No matter where an airline is in its flight-efficiency and sustainability program, Honeywell can help it take the next major step — and the step after that — towards a smaller carbon footprint and simplified tracking and reporting, thus engaging the entire organization to promote a fuel and carbon reduction culture. Additionally, it can help optimize current best practices, unearth new fuel-saving opportunities, or perform a deeper analysis of operational data in support of the broader sustainability commitments of the enterprise,” said Bob Buddecke, president, Honeywell Connected Aerospace.
The program’s decision comes from analysis of the variation between the planned and actual fuel used over hundreds of flights between city pairs and a two-year period. It is easily integrated with major airline flight-planning systems such as LIDO and Sabre DM/FPM, while easy-to-read data displays let flight crews and dispatcher compare options and clearly see the impact of their decisions on fuel consumption. Data also is normalized to reflect seasonal variations, like changing weather patterns that can cause delays or diversions.
To ensure consistent accuracy, Honeywell continuously monitors the custom algorithm for each operator to ensure that it accurately reflects the fuel-loading recommendation for each city pair flown.
StorkJet, based in Katowice in Poland, helps airlines save fuel with advanced data analysis. Thanks to artificial intelligence they monitor precisely aircraft performance and optimize fuel consumption across 44 fuel initiatives. Its customer base includes JetSMART and Volaris in the South America and Air Atlanta, LOT Polish Airlines and Wizz Air in Europe. Air Astana signed up in July 2021.
Piotr Niedziela, co-founder and head of Business Development, explains that one of the company’s products is AdvancedAPM which helps airlines precisely monitor aircraft performance and diagnose root causes of performance degradation. It compares actual fuel consumption of each aircraft to a brand new one. The difference is called performance factor and is later on used in multiple systems like FMS and Flight Planning System to properly plan fuel for the flight, as well as optimize speeds, altitudes and trajectory of the flight.
The second StorkJet product is FuelPro — a fuel efficiency platform with over 44 options across the entire spectrum of airline operations which can be optimized, including fuel policy, flight planning, ground operations, APUs, departure, flight speed, vertical profile optimisation and arrival. Each fuel initiative has a dedicated AI mode, that creates ‘what if scenarios’, compares fuel burn between them and indicate the precise savings potential that an airline can achieve. Such powerful analysis can be further broken down into time periods, airports, runway, aircraft type, individual aircraft and many more.
As pilots are the most important factor in the efficiency process, with FuelPro they also have dedicated app where they can check their fuel score. They are being encouraged in a friendly way, being informed how much CO2 or trees they have saved with efficient flights. After each flight they receive a debriefing with information on how close they were to optimum policies, like optimal flight speed and vertical profile during climb, cruise, and descent.
One example of where fuel savings can be made is the amount of additional fuel that is carried. There are many different types of additional fuel which could be safely optimized like contingency fuel, final reserve, discretionary, alternate fuel etc. Looking at the numbers, even with 100kg fuel less onboard at an airline operating 50 aircraft can save between $120,000-200,000 annually. See Table 1, previous page.
In the case of LOT Polish Airlines, it used FuelPro to reduce Contingency Fuel on their Boeing 787 fleet. The result was a change in policy from 5% to 3%, which brought savings of $1.8 millions in just one year.
With FuelPro airlines can also easily check what is the real saving potential with using low acceleration altitude and low flaps during take-off. What might be interesting is that with AI models airlines can check different scenarios and be aware of real impact of non-compliant flights. Averages, based on 1 mlnute of flights are shown in Table 2, previous page.
At the moment the company is working on an EU-funded project for real time optimizations — fuel briefing and debriefing via the EFB.