In 1984, John Zachman, then with IBM, proposed the Enterprise Architecture (EA) framework. Though now upgraded to Version 3.0, the core ‘5W1H’ (Who, What, Where, When, Why, and How) EA structure remains unchanged. At an atomic level, the EA framework focuses on enabling the functions, processes, management rules, and Office Automation (OA) work to meet the requirements of IT-powered businesses.
The EA framework has given birth to a wide range of commercial software platforms — such as Enterprise Resource Planning (ERP), Product Lifecycle Management (PLM), Customer Relationship Management (CRM), Enterprise Asset Management (EAM), Manufacturing Execution Systems (MESs) — that are the backbone of modern business processes.
To date, digital enterprises have evolved through three important phases:
The 1970s and 1980s were the era when mainframes were the most capable tools to handle complex computing tasks.
The 1990s were the time of informatization, with the aim to solve information processing and decision-making tasks for service management.
The Internet era, phase three, came into full force at the beginning of this century and is concerned with solving the cross-regional problems encountered in eCommerce and communicating in multiple languages.
Through the first and second decades of the 21st century, Germany, the United States, and China have launched the Industry 4.0, Industrial Internet, and Made in China 2025 initiatives, respectively, signifying a new era of global intelligence. The early stages of this period could be characterized as the sharing phase. The challenge of the sharing phase is to solve behavior digitization to enable smarter decision-making. The next, higher stage will be the intelligence era — which will require us to reconsider and rebuild the entire enterprise architecture. To do this, Huawei has introduced the Smart Enterprise Architecture Framework (SEAF).
Building a Contextual Model
In the current sharing phase, organizations are focused on cloud computing, the Internet of Things (IoT), and Big Data in a move towards digital transformation. This change is necessitating a shift in focus from digitizing functions to a new model for capturing the behaviors and interactions between people, things, and events.
To answer this question, Huawei has built a Smart Contextual Model (SCM) with the core purpose of supporting the digitization of people, things, and events using high-performance processing and full connectivity as key technologies.
Smart contextual model
Digitalization of People
Informatization is mainly focused on the results of people’s operations on events (business flows) and things (their use).
The behaviors of people include interactions with enterprise, virtual, public, and private spaces. By focusing on individual behaviors in each of these different spaces, it is possible to capture a complete digital experience.
The behaviors of people who are members of an enterprise are expected to align with the governance, mission, culture, and capabilities of that company. Digitization leadership means that the mindset and skills of managers must be captured and encoded. IDC, the research and analysis company, has reached the conclusion that 33 percent of leadership-level staff must be profiled accordingly.
Digitalization focuses on the actions and results of people in different situations with the goal to analyzing peoples’ behaviors for improving business processes to improve efficiency and create more value for the enterprise.
Digitalization of Things
The ability to manage ‘things’ relies on a combination of external interactions and internal operating behaviors.
21st century Cyber-Physical Systems (CPS) — including devices, tools, and production and logistics equipment — are expected to digitize their internal operating behaviors and connect with the outside world. Therefore, making things smart and aware requires comprehensive models of all internal actions and external interactions.
The mapping of behavioral relationships between people and things across time and space will achieve unprecedented levels of integration and collaboration. By using Big Data analytics, we expect to discover countless dependencies that will be used to improve government, economic, social, educational, and healthcare outcomes.
Digitalization of Events
Data enablement and the informatization of service functions are focused on the results of value creation during business processes.
With customers at the center, business organizations are restructured to transform department-centric processes to role-based processes. The result is the creation of value E2E through the connection of relationships between people, locations, things, and time.
Business processes are rearranged using a service-centric approach by moving toward automation and intelligence. Business process results are analyzed in order to break data barriers and optimize E2E data sharing. The value creation process is improved by increasing E2E efficiency and customer satisfaction. Business-model and service innovations drive the transformation of operation models.
Two key technologies are critical to complete all forthcoming SEAF implementations:
Enterprises and organizations must create a detailed framework for data governance, to promote enterprise transformation and drive business-model, management, product, and service innovations.
Enterprise data is divided into structured and unstructured data. Unstructured data is further divided into repetitive and non-repetitive data. The effective use of data for event handling and decision-making requires architecture and relationships construction, behavior modeling, and smart analysis.
In different digitalization stages, models are divided into three categories: 1) Entity relationship models meet the requirements to implement the business logic for Online Transaction Processing (OLTP). Advanced frameworks must also include behavioral entities as a part of the model. 2) Dimensional Fact Models (DFMs) are necessary to support the conceptual representation of the Data Warehouses (DWs) used for smart business analysis. 3) Big Data correlation models based on 4V features (Volume, Variety, Velocity, and Veracity) are focused on the production and value-mining requirements that enable the unified modeling of facts, dimensions, and behaviors.
Technologies that enable digitalization are divided into three categories: 1) Smart devices: Including sensors, Customer Premises Equipment (CPE), IoT terminals, and other end-point technologies. 2) ICT hardware: Including smart terminals, network switching and transmission facilities, and cloud computing and storage facilities. 3) Software: divided into 6 categories: embedded, edge computing, business services, analytics, platform services, and resource layers.
White-box ICT hardware and software-defined hardware are indicative of two trends: 1) Software-Defined Everything (SDX) is inevitable. Software is the core of the digitalization era, including all forms of collection, transmission, processing, and storage. 2) Software is the primary tool for solving many of the core problems facing most enterprise processing platforms, including timeliness, accuracy, hands-off automation, intelligence, and volume.
Connectivity technologies must link people, things, and events. Connectivity has given birth to the Internet, mobile Internet, the IoT, and finally an Internet of People (IoP).
SCM outlines a comprehensive digitalization methodology that leverages all six domains of the original EA framework — applications, data, technology, time, people, and organizational motivation — plus three more domains added by Huawei: business processes, things, and integration — for a total of nine domains. These domains enable full connectivity and digitalization, and are summarized below in contextual order.
Business processes that create value are the foundation for enterprise strategies. In turn, the visualization of E2E service capabilities is the operating mechanism of each business model design. Of course, there are also the core objects of digitalization and informatization: business processes must be customer-centric, role-based, and focused on service value and organizational behavior.
Customer-centricity is a core philosophy that fosters service-oriented cultures. The introduction of service- and role-centric processes requires changing away from antiquated practices that were most concerned with departments and functional entities.
Hierarchical and structured thinking establish a complete business process architecture. Responsibility mechanisms ensure that backbone processes are simple and stable, and low-level processes are flexible and standards-compliant. Streamlined backbone processes across the value chain achieve improved E2E efficiencies and customer-to-customer integrations.
Integrated business model processes provide internal control, authorization, monitoring, and other decision-making behaviors. Business processes must support the ability to trace the organizational record and individual people’s behavior.
Service applications are the natural result of a complete business process architecture. The purpose of applications is to abstract business activities into the smallest possible number of functions and operations based on service boundaries, data associations, security, and many other factors.
In the informatization stage, data models are constructed based on the inputs and outputs of the business processes. Structured data will include transactions, business rules, and the interrelationships between reference data. Dimensional models are built on business and management indicators, process performance indicators, and other management and decision-making criteria.
In the digitalization stage, the focus is on objects and their behavior for the purpose of constructing models based on business behavior, peoples’ behavior, and the interactions between things and things, and people and things.
Unstructured data (repetitive and non-repetitive) can be used to establish relationships between people, things, and events. Big Data analytics can be used to cover scenario- and behavior-specific trends and correlations in the areas of customer churn, user loyalty, user habits, disease monitoring, emotion analysis, claims, fraud, violations, location services, biometrics, and many others.
The integration of business processes, service applications, and data, among others, also involves vertical, horizontal, and E2E integrations based on industry chains and smart manufacturing. Additionally, digital transformation calls for the integration of the models dedicated to the aggregation of data from people, incidents, and things.
Considering the universality and scaling characteristics of ICT technologies, it is important to plan and develop an Enterprise Technology Framework (ETF) to cover policies, architectural standards and specifications, Application Programming Interfaces (APIs), reference models, and so on.
Technology sharing platforms based on ETF are best constructed to include networking, cloud computing and storage, and the IoT for heterogeneous integration and unified access to application portals. At the same time, security platforms need be constructed based on cloud-pipe-device and data components.
The integration of platform-based technologies is the foundation for the business strategies necessary for fast responses to globalization.
Time is a special case for digitalization and informatization, as IT applications manage time differently through each stage of the business process.
In the digitalization stage, Big Data analytics are used to track time stamps and value trends. Business modeling based on inputs from people, things, and events can be organized in time sequences to discover trends through analysis. Such modeling makes it possible to predict upcoming events in order to innovate new business processes and provide more refined services.
In the informatization stage, business processes track the time stamps of key service events. Key time points are tracked throughout the business process in order to better manage the efficiency of creating and modifying products and services.
In the R&D phase for designing intelligent things a wide range of possible characteristics need be taken into account. Behavioral models may include internal- and external-facing sensors, connectivity requirements, self-learning, and self-healing depending on the use-case for each type of device. Whether for machine-to-machine interconnection, or interaction with people, every thing on the network must be integrated with a management application system for operations and control.
Socially, people display different roles based on context and circumstance. In professional environments, the organizations that adopt EA frameworks have to construct space-behavior models that define roles based on value creation and management control requirements. The organizations that construct effective behavior analysis models benefit from efficient business processes and compliance with internal controls.
Role-based approaches optimize the distribution of resources based on organizational structures.
Digitalization and informatization continue to enable revolutions in management, streamlining of business processes, and the acceleration of enterprise transformation. History proves that digitization and informatization remain a top priority for enterprises and organizations that are committed to fulfilling their strategic visions.
Digitalization and informatization technologies are key elements for energizing business change and transformation. IT departments must work closely with the business strategy departments to promote effective leadership.
Architecture Planning Guide
SEAF is an atomic-level framework that defines the hierarchical components and key elements within each domain. Building smart enterprise architectures for digital transformation benefit from design methodologies such as The Open Group Architecture Framework (TOGAF).
TOGAF-based High-Level Designs (HLDs) include the following three aspects in order to protect long-term investment, implement and operate transformation plans, and continuously improve and optimize these plans:
Digital transformation strategy planning includes organizational and operations planning, cloud, connectivity, and Big Data strategies.
Top-level designs for digital governance include overall enterprise strategies such as vision, mission, first-order principles, market position, organizational control processes, resource distribution, long-term investment, and performance and incentive measures.
In short, if enterprises want to complete their digital transformation, they must do the following three things.
First, transform the philosophy and ideology of management. Management leaders must pioneer and promote the benefits of digital transformation.
Second, management must employ TOGAF and HLD-based best practices to adhere to a good implementation blueprint.
Third, prepare well for each aspect of digital transformation, including organization, capabilities, capital, and so on. Conduct plenty of research and verification on the eve of transformation, widely agree on and recognize digital transformation challenges, and prepare for long-term investment and perseverance.
(Suggestions and comments from enterprise architecture experts regarding this article are welcome and encouraged.)