Today, every interaction, every system, every touchpoint, and every sensor generate vast quantities of data. And there are very few industries that generate more of it than aviation.
The average commercial aircraft can generate as much as 20 terabytes of data per engine for every hour that it’s in flight. Multiply that by the total engine hours accumulated across an international fleet, add in a myriad of other passenger and airport data, and it’s easy to see why managing and getting maximum value from data have become such significant challenges for today’s airlines.
From detailed performance information to valuable passenger insights, aviation companies understand the value trapped within those vast volumes of data. However, with such a wide variety of sources to manage, and new capabilities needed to translate them into actionable insights, realizing that value isn’t easy.
That’s where Data Mesh comes in. Data Mesh is an analytical data architecture and operating model where data is treated as a product and owned by the teams closest to it. Instead of a single, central team handling every data-related request and managing everything, different domains are responsible for managing their own data products. A bare minimum is centralized to enable safe use and easy access to data across an organization, while empowering domain teams to innovate locally, however they choose.
It’s quickly rising in popularity in industries where organizations must manage and integrate data across diverse domains — just like aviation. Here’s a look at how the architecture approach is helping airlines and aircraft manufacturers tackle five of their biggest and most persistent data challenges.
#1) Connecting with customers and differentiating through distribution
When the IATA’s New Distribution Capability was launched about 10 years ago, it began to transform how air travel services are offered to customers, giving airlines a new way to differentiate and personalize their offerings, while improving choice and visibility for passengers. But with that shift came significant challenges for airlines, and players across the air travel value chain.
Airlines had to learn a lot about their customers very quickly to understand how, where, when, and why they buy travel services, and build compelling offers that meet modern passenger needs. To do that, they had to gather, process, and operationalize customer data across numerous systems and silos — a challenge many are still grappling with today.
Using Data Mesh, the teams closest to customers that best understand their changing needs can become the primary custodians of customer data. That enables an airline’s customer and distribution teams to plan and deliver their own data initiatives from end-to-end, helping them become more responsive and empowers them to achieve their strategic goals with minimal outside support.
In practice, that means being able to do things like personalize offers to drive revenue, and build high-value loyalty schemes that don’t just support customer retention, but generate insight to enable the continuous improvement and optimization of airline offerings.
#2) Breaking down data silos across domains
Aircraft data. Passenger data. Distribution data. Airport and ground operations data. They’re very different, but all equally critical to aviation operations. With a huge range of systems, touchpoints, and architectures in use across those domains, for many airlines those data types are highly siloed, making them hard for teams to explore and operationalize.
When poor weather or mechanical issues strike, everyone one of those vast and disparate data sources need to be brought together to help solve the issue and get operations back on track. Customer and scheduling data is needed to rebook and amend flight plans. Aircraft and crew data is needed to ensure that it’s safe and legal to fly. And airport and external data are required to request new takeoff and landing slots, and make sure that conditions will be suitable to fly in.
Data Meshes are built for scenarios like that. In a Data Mesh, each domain’s data is owned and managed by that domain’s experts. They’re the ones that know that data best, so they can use their expert knowledge to define what high quality, fit-for-purpose data looks like for them — building data products that meet their needs.
Crucially, however, every product each domain team creates can easily be made visible and accessible to every other domain team too. So, when a crucial situation arises like a technical fault grounding an aircraft, it’s easy for teams to bring disparate data products together and respond quickly and effectively, minimizing disruption.
The end result is a data architecture where there is clarity and distinction between business domains, but the walls between data silos are broken down. Everything is highly discoverable, but still owned and controlled by the teams closest to it, helping to support data innovation.
#3) Streamlining international data governance and compliance
With operations spanning multiple continents, and a huge range of regulatory requirements to meet, data governance and compliance are especially challenging for aviation organizations.
In Data Mesh architectures, governance is both decentralized to each business domain and federated for coordination, supported by the platform that sits below each data product. This helps every domain meet specific compliance needs quickly and efficiently. The data generated by that platform can even be exposed as its own compliance product. So, when a regulator needs access to an airline’s data, or requires proof of how data is being managed and used across borders, it’s immediately available.
Plus, because Data Mesh architectures enable federated governance and secure collaboration on data products, airlines can even invite regulators to collaborate on their data governance products. In doing so, they can go above and beyond what’s demanded — exceeding transparency expectations, and building customer and regulator trust.
#4) Multiplying the value of data with Artificial Intelligence
From dynamically pricing tickets and services, to determining the right levels of fuel to use for every flight, today’s airlines use data science and machine learning models to make many important decisions.
Those models automatically translate the data generated by airlines and aircraft into insights, enabling routine decisions to be fully automated while also empowering human teams with valuable insights into patterns and trends without needing to manually dive into data and process it themselves.
Data Mesh provides the foundation for AI success by enabling teams to build solutions based on timely, high-quality data that the entire organization can trust. This enables your data scientists and ML teams to spend more time solving valuable business problems and less solving avoidable data problems. Teams start by determining what they want a model to do, then build a data product that pulls in relevant data sets for training. At that point, they can apply automation to enable models to continuously learn and drive business value automatically.
#5) Building cultures of experimentation and innovation
The COVID-19 pandemic has driven organizations across the aviation industry to accelerate innovation and improve their operations. Fewer business customers are traveling, international flight demand remains suppressed, and operating costs are rising — driving airlines to transform what they offer, how their offerings are priced, and who they offer them to.
Airlines and other aviation companies haven’t had the luxury of being able to tackle those challenges in isolation either. Ongoing shifts such as the drive to reduce emissions and increase the sustainability of air travel have not only continued, but accelerated — giving companies a huge amount of change to think about, respond to, and proactively lead.
To succeed in these shifting conditions, airlines need the freedom and capability to experiment with new data-driven services and offerings, learn from their experiments, and quickly scale up winning concepts that demonstrate strong ROI.
By making expertly-curated data products highly discoverable and giving every team the power to build their own data products, Data Mesh enables organizations to experiment with their data and operationalize it in new ways.
So, if a distribution team wanted to experiment with a new, highly personalized way of targeting and engaging specific passengers to drive loyalty and revenue, they have the power to do exactly that. Then, if their experiments prove successful, the entire business can see the value in scaling those experiments out and using them to drive a profitable and sustainable future.
Discover what Data Mesh could do for you
Those are just a few examples of how Data Mesh architectures can help organizations in the aviation industry tackle their data challenges, break down silos, and start driving greater business value from their data.
Data Mesh offers many benefits, but canny readers will recognise that it is not simply a technology solution. We have found that courageous leadership, business engagement, a willingness and ability to transform how teams are organized, and disciplined software engineering practices are all factors contributing to successful adoptions.