Disruption need not bring misery

Flight cancellations and delays are painful for travellers and airlines alike. Passenger-centric tools, however, promise to alleviate woes by creating positive opportunities for both sides

When you are one of hundreds of passengers whose flights have been delayed or cancelled, it is a dispiriting and frustrating experience. More likely than not, you will be standing in line waiting for information, or vouchers for a meal or hotel. Never mind those plans for business meetings, family reunions or holidays that are now in disarray.

Disruptions also put airlines under pressure as they work to get their passengers' journeys back on track. This can take time, during which the costs and stress rack up for all involved.

Understanding the full impact and cost of disruption is not easy. Some costs, for example staff overtime, are easier to quantify than indirect costs, such as negative exposure on social media.

Amadeus and T2RL estimate the cost to carriers as up to 8% of worldwide airline revenues, which amounted to some $60 billion, according to their 2016 report, Shaping the Future of Airline Disruption Management. Using IATA's global airline revenue projection of $824 billion in 2018, we may be looking at around $66 billion being wiped off the bottom line by disruptions this year.

But what if these situations could be turned on their head by progressively managing the event, rather than reacting after it? Machine learning, predictive technologies and automated tools are emerging that will enable airlines to take a holistic approach to optimisation and recovery, allowing trips and schedules to be resumed more rapidly.

These tools will also help airlines position their customers at the heart of the solution. Better still, this passenger-centric approach offers opportunities to turn your response into a brand differentiator and generate ancillary revenue.

These solutions, however, are more than a matter of adopting new technologies, assimilating data and integration into existing systems. These solutions will require a total realignment of mindsets and the way the industry works.

More than 400,000 flights were cancelled across the world last year, with some 9% of services arriving more than 30 minutes late
More than 400,000 flights were cancelled across the world last year, with some 9% of services arriving more than 30 minutes late

Herbert Kehrer/imageBROKER/REX/Shutterstock

A comprehensive view of operations that utilises predictive tools to plan resource management will be a major step forward in tackling disruptions, according to Sebastien Fabre, vice-president of airports at SITA. "Today, most of our industry's resources are planned and deployed based on carriers' published flight schedule. With an industry-wide on-time performance of around 76%, this is an area where more accurate data could dramatically improve efficiency through proactive planning and management. This will directly improve the passenger experience of nearly one flight in every four," he says.

SITA has worked with Constraint Technologies International to create a real-time, event-driven situational awareness tool that aggregates data from core airline systems to provide a comprehensive view of the fleet, passengers, maintenance, crew and cargo on the day of operation.

After a successful proof-of-concept trial with a European airline in 2017, SITA will be officially launching FlightAssist in early 2018. The tool is designed for mobile devices and can share information across a much wider group of stakeholders in the airline and at the airport – not just those running the operational and hub control centres.

PREDICTING CONSEQUENCES 

Meanwhile, artificial intelligence capabilities are being developed by SITA Lab to better predict delays before they have an impact. Fabre envisages that these, plus CTI's "Solver" technology, which will allow "what if" scenarios to be run, will both be integrated into FlightAssist in the next 12-18 months. "While it is impossible to predict certain events, we can better manage their impact by predicting and proactively managing the consequences in the air transport industry," he says.

GE Aviation is also working to give airlines a more holistic view of disruptions. It has deployed its Predix Industrial Internet platform and machine learning techniques for more than 10 years to improve engine reliability and troubleshoot fleet issues that impact operations. However, chief technology officer Jon Dunsdon suggests that innovation developed for engine optimisation could also be harnessed for disruptions. He cites GE's Digital Twin Technology, which provides a view of the past, present and future of a critical industrial asset such as a power-generation plant.

"We see the Digital Twin as a key technology to fully digitise the physical world and interact with it using 3D augmented reality headsets and voice recognition. Some of this past, present and future view is being applied to assist the airlines' operations decision-makers in planning more robust networks, as well as in dealing with irregular operations – IROPS – as they occur," says Dunsdon.

GE IROPS recovery optimisation has been used for 15 years to reschedule flights and crew. However, in future Al will help optimise all resources, right down to creating alternative travel itineraries suited to the individual passenger, says Dunsdon. "AI brings an affordable capability to continuously learn from big data as airlines operate and fine-tune the optimisation to changing operational and business demands."

Automation will be key to helping airlines cut through disruption issues in minutes, as opposed to hours. Lufthansa Group airline Swiss has ramped up the speed of rebooking its passengers on to alternative flights during disruptions by replacing manual processes with automated solutions that harness industry-wide data. Swiss collaborated with Amadeus to deploy a centralised and automated disruption solution linked to Amadeus's Altea system and interline partners.

Freedbird's app enables business travellers to book an alternative US domestic flight
Freedbird's app enables business travellers to book an alternative US domestic flight

Freebird

Swiss played a major role in the design and development of the tool from a business perspective, including calibrating the entries of business rules so that the tool automatically finds the same or better solutions compared with a manual process. The entries and processes were designed so that Lufthansa Group partners Austrian Airlines and Lufthansa could also use it without having to develop their own airline-specific entries.

"The biggest advantage is the speed of the rebooking, which means that the guests get an alternative connection much faster than before," says airline spokesperson Stefan Vasic. "For instance: a long-haul flight with approximately 250 persons is analysed and rebooked completed within approximately 10-15 minutes." Customer-facing staff have more time for personalised passenger care, and Vasic notes: "With perfect handling and a fast information chain, an airline can distinguish from its competitors in irregular [operations] cases."

Overall, consumers' expectations of their airline during flight disruptions are increasing. "Every day, travellers are coming more and more to expect new mobile technology, elegant user experience, and seamless real-time data in every part of their lives – whether they're calling a ride (disrupted by Uber), trying on new shoes (disrupted by Zappos), or buying home supplies (disrupted by Amazon). Why should air travel be any different?" asks Ethan Bernstein, chief executive of Freebird.

EMPOWERING TRAVELLERS 

Freebird is one of the AI travel assistants that have emerged over the past couple of years to empower primarily business travellers during disruptions. The mobile solution allows US domestic passengers to skip the queue and instantly book a new ticket – on any airline at no extra cost – in the event of a flight cancellation, significant delay or missed connection.

Since its launch in 2015, Freebird has gained traction in the corporate travel market, partnering with 10 agencies representing $7 billion or more of "downstream client spend", integrating with the three GDS reservation systems, and being used by dozens of corporate clients.

Travellers can purchase Freebird any time after their itinerary is confirmed up until about two days prior to departure. Fees are determined by the app's proprietary algorithm, which predicts the risk of disruptions for millions of flights per year. It also provides users with instant notifications in the event of a flight disruption and offers the ability to rebook with only three taps on a mobile device.

Bernstein acknowledges that airlines have made remarkable progress over the past few years, providing apps to accommodate passengers, tracking aircraft in transit and providing push notification alerts, but believes there is room for improvement.

"A little over half of Freebird rebookings are made on other airlines, implying that the airline re-accommodation was not attractive relative to other availability in the reservation system," he states. "Said differently, nearly half of Freebird rebookings are made on the same airline the passenger was already flying on.

New platforms can help airlines and passengers find hotel rooms when problems occur
New platforms can help airlines and passengers find hotel rooms when problems occur

Mike Stocker/Sun Sentinel via ZUMA Wire/REX/Shutterstock

"Clearly those passengers didn't even get the best flight available on their existing airline. We believe putting power back in the hands of the traveller is a more efficient way to solve problems, give the traveller a feeling of control over their situation, and ensure positive outcomes."

Moves are under way in the airline community to provide passengers with direct control and instant choices. Travelport Resolve, launched in July 2017 after a trial with a major US airline, automates the sourcing and distribution of hotel rooms, air rebooking, plus meal and transport needs in a single, seamless – and paperless – experience.

"The whole idea for Resolve is to take as many people out of line as possible, so that airline personnel can concentrate on the hard issues," says Mike Melton, Travelport's senior director of technology. "We will be continually looking for ways to meet this goal and provide a one-stop shop for disruptions. As part of that one-stop shop we are looking at [a] chatbot based on AI and natural language processing to give the passenger as much input and communication as possible."

He believes that in the near future the industry will automate everything for the passenger in the disruption process, based on the passenger's status, personal information, itinerary and issues. "Predictive analytics and AI are going to help forecast which hotels are likely to be selected, how many rooms are going to be needed, what is the best itinerary for a passenger based on their input and needs, and what alternatives are available. Smart chatbots are going to be able to 'communicate' with passengers. In short, the disruption process is going to become more personal," says Melton.

DISCOUNTED RATES 

Justin Steele, vice-president for product at travel commerce platform Switchfly, agrees that passengers have to be at the heart of the solution. Switchfly launched its own IROP Management solution in August 2017 to give stranded passengers hotel choices and give airlines access to deeply discounted rates. This hooks into airline departure control systems and back-end functionality. Business rules allow airlines to specify the service offers for their passengers. However, the true upside of the disruption situation for Steele is the opportunity to generate ancillary revenue.

"I am really excited by the ability to increase revenue, which will help offset some of those hard costs as well. For some reason when we are giving customers something for free, we are afraid to up-sell," he says, explaining that there is also scope to give passengers a chance to upgrade the basic, free, offer.

Gaining passengers' trust is central to developing a retail opportunity and this starts by offering choices. "Some individuals might be there on a leisure trip and they might want to go back to the downtown – they want to have a nice dinner and they want to have the option to extend their vacation. Another business traveller might be backed up on email – this is a huge inconvenience and they want to be as close to the airport as possible."

Recognising the passenger's frequent flyer status is also critical: "We treat our top-tier customers within the industry extremely well, but it seems like any time there's an irregular operation all that goes out of the window and it's just a panic. By identifying, 'yes I understand you are one of my top tier members, and so I am going to give you a little bit extra spend to go get your accommodation' can go really a long way," says Steele.

And the final key to unlocking the ancillary revenue potential is allowing passengers to self-service via their mobile device. "Once you have gained the trust of your customers and they understand that you are really looking out for their interests, there's a whole world of retailing opportunities."

Queuing to rebook flights could become unusual as the need for human intervention falls
Queuing to rebook flights could become unusual as the need for human intervention falls

Xinhua/REX/Shutterstock

Data tools will also help airlines gain more insight into the cost of disruptions and avoid wastage. Switchfly is able to do live pricing calls to determine inventory at the time of the disruption and help the airline manage the rate. Steele says: "There's already some big data we're looking at – what's the fair market price in that market for that specific night and how much inventory can I get there?"

Melton explains Resolve can track the reason for the disruption, the hotels that are being used, the costs, the time of day, and even who booked through which channel and provide it to the airline in real time. "Many airlines have to wait months to see this data – if they get it at all. Analysis of this data can help determine if the right people are getting rooms, if their rules are too liberal or stringent, if they should focus on ways to communicate or a certain channel, what hotels are their best partners," he says.

Automation of disruption tools has also percolated down to the claims process. Historically these could take up to 21 days after a disruption to work through. However, Intelenet Global Services has cut that down to roughly one day with its iCAN tool.

ELIGIBLE CLAIMS 

Travel and logistics sector leader Bill Hoppe says: "Depending on the amount of information the airline is willing to share with us proactively, we can categorise a lot of these claims in advance, so before the claim comes in we would know if there are eligible claims for a particular flight or issue."

He adds that 40-50% of these claims do not need to be reviewed by humans, apart from basic quality checks. The machine learning AI technology means the tool will recognise when something does not make sense or is outside parameters and refer it to an agent. The agent provides more information to the tool so it learns and the resolution rate improves.

Hoppe forecasts that in the next year or two, 80% or more of claims will need no human intervention. "You could get to a point where you are proactively offering compensation in a lot of situations. We don't yet have that level of automation with any of our customers, [but] some are approaching that."

Automation and AI tools demand consistency and this is a stumbling block for creating a seamless passenger-centric disruption strategy. Hoppe says policies and procedures around claims for IROPS events are inconsistent. "I use the example of some of the hurricanes that hit the Americas in the last year – just the wide degree of variability in process and procedure for how those were handled, what flights were available for compensation, which were not. It can vary by a matter of minutes and can change constantly throughout the day, so tracking can be extremely challenging."

Steele adds that within a single airline there can be different agreements, rules and regulations across different locations, which has created a disjointed user experience for passengers. "Trying to have a unique experience, even within all the brands in your alliance, is going to be a strong step forward."

New technologies are only half the battle to improve the industry's ability to recover from disruptions. Implementation will demand new processes and fresh approaches from airline teams, managers and stakeholders as they become proactive, rather than reactive. Fabre says this will be the biggest change of all.

"It will require many of the processes within the industry to be re-engineered to support a proactive management of a fluid network and ground-based operation," he says. "Most importantly, this change in management approach must also address 'work practice change management', as an initiative of this type will fundamentally change the way in which team members will interact with one another and their stakeholders... driving the need for greater collaboration."

Switchfly offers new revenue opportunities
Switchfly offers new revenue opportunities

Switchfly

Airline disruption trends

30% of airlines accurately capture the full cost impact of delays – 57% have plans to do so in the next three years

Re-accommodating passengers when disruption and cancellations are unavoidable is the most important benefit technology can provide for 69% of airlines

Improving data quality and analysis is the most important disruption management priority for 72%

99% are planning to invest in real-time information and alerts via passengers’ mobiles over the next three years

82% are planning funding for self-service solutions for irregular operations

80% are set to pump funds into disruption planning/predictions systems

Source: SITA Airline IT Trends 2017 

www.sita.aero/insights-2017

Flight disruptions in 2017

35,714,925 flights operated

401,958 flights were cancelled

5,435,186 (15%) flights had a delay to arrival of more than 15 minutes

3,088,656 (9%) flights had a delay to arrival of more than 30 minutes

Source: FlightGlobal, incorporating FlightStats, based on data from 99% of scheduled flights