• banner pc

    Streamlining Data and Services Empowers DX

Enterprises in the transportation industry pursue digital transformation so that they can more effectively manage and control their service flows. With this in mind, the priority in helping transportation enterprises to thrive by digitally transforming becomes clear: We have to enable them to streamline their connected data and services.

The aim for enterprises in the transportation industry is to optimize service flows to enhance safety, increase efficiency, and improve the services they provide. This is their primary concern and is central to their digital transformation.

Streamlining Two Focuses

To successfully digitally transform, digital upgrade is necessary. It helps build a data exchange platform to centrally collect and store data that is scattered across data silos and is isolated.

Full connectivity and refined management and control of service flows can't be achieved with numerous data silos. While eliminating them doesn't guarantee success, it's the first and most fundamental step according to many failure cases.

To meet transportation enterprises' service needs for digital transformation, we need to address two issues: connecting data and connecting services.

Connecting data

Digital upgrade must be driven by the value of data. To ensure that happens, there are several questions to consider: How can data value be maximized? Does data connection mean to collect scattered data from various service systems and centrally store it? How can we implement data governance? Should we name data in a unified way or convert it into a unified format?

Connecting services

To an extent, digital upgrade is a process of redefining service processes without changing departments' functions and responsibilities. For example, digital upgrade could involve determining how airports can manage and coordinate all processes based on service flows.

Only by streamlining the two focuses can we truly empower enterprises to successfully digitally transform. First and foremost, a new data-driven service mode needs to be built — to fully connect data and services, capitalizing on the value of data flows.

IOC: Service Processes

Digital transformation typically involves the creation of an Intelligent Operation Center (IOC), which aims to enable global dynamic visualization against complex service flows and implement situational awareness, service collaboration, and assisted decision-making.

For example, flight operation support — the most critical airport services process — includes taxiing guidance, stand allocation, and ground handling for inbound flights as well as the waiting, taxiing, and takeoff processes for outbound flights, and it involves multiple units, business departments, and work procedures (personnel, tools, steps, and so on).

This process must ensure safe and efficient arrivals, ground services, and departures, and it must meet airlines' needs for transporting passengers and carrying cargos. Accordingly, an airport usually develops multiple information systems — such as a ground handling system, a Flight Information Display System (FIDS), a Departure Control System (DCS), and an integration system — to support its services. The Airport Operation Center (AOC) needs to handle multiple systems and terminals: It organizes airport staff to analyze and compare various information sources and deliver dispatch and command instructions through handheld terminals.

When using multiple service systems for a single process, a lack of End-to-End (E2E) management and control of service flows makes it difficult to enable collaboration and scheduling within seconds in complex scenarios. This is why customers require digital upgrade.

Scenario-based solutions for data-driven services, instead of data exchange and integration, are needed. The IOC connects services through two-layer modeling — service modeling and data modeling.

Service modeling aims to achieve the objectives and address challenges of management and control of service flows, monitor all operations along the service chain to better implement more effective collaboration and scheduling across various tasks, departments, and units — ensuring safer and more efficient flight operations. Service modeling raises new requirements, such as situational awareness, surface monitoring, coordinated scheduling, and decision-making support.

The IOC constructs service flow models based on a service overview. It needs to answer the following questions: What are the users, scenarios, requirements, and problems? What information should be obtained? What commands need to be delivered?

Take the air route view, designed for resource scheduling agents, as an example. It sorts and tracks inbound flights and focuses on the agents' biggest priority — airport resource allocation.

It then addresses questions such as: Can the system dynamically calculate a more accurate Estimated Time of Arrival (ETA) based on flying and approaching status? Can the system predict the stand arrival time based on surface operations? Can the system predict potential stand conflicts based on the aircraft stand arrival time and rules of stand usage by preceding and following flights? Can the system automatically devise a more reasonable stand adjustment plan in the event of stand conflict risks? Once the plan is determined, can the system adjust the stands in one click and synchronize the adjustment to all operation systems according to service rules?

The real-time terminal passing status, support process of preceding flights, and stand capacity are displayed on the air route view, to provide comprehensive airport running status for dispatchers — facilitating collaboration and scheduling.

Based on service modeling, an E2E mode is established — helping airports implement sensing, management and control, collaboration, and decision-making support of the entire service chain by pre-event forecast, in-event collaboration, and post-event analysis.

Undoubtedly, data plays a critical role in E2E management and control of service flows, so data modeling is crucial. Data modeling doesn't mean to exchange or integrate operation data of multiple service systems, or reconstruct the original system functions. Instead, it focuses on bridging data and process breakpoints and uses model algorithms and indicator systems to better manage and control the digitalization of processes and rules.

By digitalizing scheduling rules, Huawei's intelligent stand allocation system uses Artificial Intelligence (AI) algorithms to automatically schedule stands based on real-time data awareness. As well as reducing manual operations, it also enables dynamic stand allocation and scheduling within seconds based on real-time surface status (including conflicts and ETA). Similarly, the Variable Taxiing Time (VTT) algorithm can predict the entire process, from an aircraft landing on the runway to moving into a designated stand. This allows the apron and ground service departments and others to perform scheduling and collaboration within seconds.

Meanwhile, Huawei's measurement indicators and indicator systems for service flow digitalization help the AOC implement in-depth management and control from operations indicators, to service indicators, and ultimately to dynamic service status.

Based on service and data modeling, the IOC serves as an integrated and collaborative service platform that fully connects all services. And the IOC platform fully supports data collection, tracking, analysis, and reports in just two weeks.

With digital transformation's influence expanding, the IOC must be continuously upgraded and iterated. It should function as both an integrated front-end platform to meet users' needs and an extensive service platform to implement functions such as forecasting, sensing, collaboration, and decision-making support. It's worth noting that building this kind of service platform doesn't mean there's any need to reconstruct existing production operation systems.

Data: Industry Data Models

Fully-connected data from service flows is a prerequisite for digital upgrade to implement collaboration and scheduling across tasks, departments, and units. To meet service digital transformation requirements, data can be fully connected by using industry data models to build a data asset platform rather than a simple data exchange, or data aggregation and report display, as a conventional data warehouse does.

When the concept of big data was proposed many years ago, it attracted huge investment from many enterprises. Those companies, including Huawei, have since conducted many practices and tried to transform their service modes to make them data-driven. All enterprises that are successfully transforming have converted data into assets that support transformation and help to generate value for enterprises by aligning with business scenarios and meeting business needs in terms of data governance, modeling, and services, as well as accumulating industry insight. In these cases, a trustworthy, available, and manageable asset platform is set up.

Huawei has developed a methodology called the "V model," which covers the process of streamlining service, data, to application flows: Sort out service capabilities by domain (sub-domain). Then, streamline scenarios, workflows, and activities. Next, work out service objects. Finally, review data entities to form conceptual models.

The key to implementing the V model into data asset construction is to build industry data models.

An overview of the industry data model

Building industry data models requires E2E governance of raw data, focusing on trustworthiness, availability, and manageability. "Trustworthiness" means that data quality must be reliable, and this is the top priority. It requires that data standards are defined in a traditional way; data is converted into assets, and business and quality standards are developed based on the business process. "Availability" refers to data service requirements proposed by businesses — a prerequisite for turning data into assets. Meanwhile, "manageability" includes data security, Operations and Maintenance (O&M), as well as operations that can support future flexible services.

The Third Normal Form (3NF) method is adopted for layered data modeling by following industry rules. Data entities are determined and their attributes are defined during the process of streamlining service flows, data flows, and data entities, forming a Conceptual Data Model (CDM). On this basis, the Logical Data Model (LDM) can be further refined based on industry standards. The 3NF, or dimensional modeling method, is used to adjust and optimize relationships between data entities — ensuring no data is missing as well as balancing redundancy and flexibility.

This modeling mode is employed in all service domains to ensure the value of data processing. Data governance, data quality, and operating flow during data processing all depend on data modeling, reflecting data mapping service process.

Service indicator systems and algorithm models are built based on the industry's standard themed library, to quickly meet data needs during the digital transformation of service processes.

Key to this is developing operations and service indicator systems to help management and business departments digitally control service processes.

Take the production and operation indicator system as an example. It defines inputs and outputs for each service activity based on the service flow, and determines measurement indicators and impact factors for service activities based on the objective of the service flow (on-time performance of flight support). Based on this, the AOC can be aware of the pressure on each support process in advance and understand the impact of the support progress on flight departure punctuality, so that it can determine whether to get involved early in the process, or coordinate with other departments, to ensure the flight departs on time.

Another priority is to focus on process and data breakpoints, to build an algorithm model warehouse and promote refined management and control of service flows. For example, intelligent stand allocation digitalizes allocation rules to enable dynamic stand allocation in minutes, greatly improving the stand turnover rate and rate of passengers using boarding gates. Aiming to enable more accurate prediction of ground taxiing time, the solution adopts the VTT algorithm to develop algorithm models based on complicated surface operations — helping the AOC implement management, control, and collaboration of apron surfaces within seconds.

At the data service layer, the data asset platform serves as an integrated platform for O&M, management, control, and operations. It enables business departments to become the owners of data asset management and service value application, truly integrating data and services.

Huawei typically sorts out service and data flows with its V model to build industry data models, implementing trustworthy, available, and manageable data governance. Aiming to manage and control service flows, Huawei develops a production and operation indicator system and an algorithm model warehouse, to empower digital transformation.

While digital transformation driven by vision brings opportunities, it also creates challenges. In the transportation industry, digital transformation means more than Information and Communications Technology (ICT) construction; it also requires digitally upgrading to data-driven service flows.

Technically, a platform-based architecture must be adopted to realize data-driven service flows, which can't be supported by a single application or functional system. In this way, digital transformation differs greatly from the previous ICT construction modes.

Digital transformation poses various challenges — in terms of technology, products, personnel, organization, and process — to Huawei, its enterprise customers, and the entire ecosystem, as it changes and optimizes conventional modes.

To address these challenges, we should focus on customer requirements for digital upgrade, for example, how data drives the management and control of service flows.

Ultimately, we can only replicate data models on a large scale — as we need to — when we build industry insight capabilities and data ecosystems.