Digital Architecture Model
Importance of digital architectures
The digitization of the physical world enables and extends business strategies with immense implications for diverse industries. Digitization influences established business models and changes the rules of the competition. There is a strong awareness of the importance of digitization in every industry. However, it is still unclear how the so-called digital transformation of enterprises should be implemented and how digital architectures differ from conventional IT environments. This section describes a model that shows the essential components of digital architectures and their relationships to each other. This model helps understand the connections and structures required for the digital transformation.
Definition of digital architectures
Digital architectures of enterprises are information technology structures in which innovative components are used that emulate human abilities such as feeling, learning, thinking, and decision-making. In digital architectures, data are generated from the physical world, transported over various networks, and stored in repositories as needed. The digital raw data are converted into applicable information that is integrated into business processes to add value to an enterprise.
Conventional IT architectures concern essential enterprise IT infrastructures (e.g. workplaces, LAN, WAN, WLAN, data centers) and IT applications (e.g. office communication, ERP, SCM, CRM, e-commerce). Digital architectures complement and extend classic IT architectures. Both traditional IT architectures and digital architectures must align to the business strategy and are intended to increase business value and enterprise performance.
Attributes of digitization and digital architectures are
- data generation by users; data collection from sensor-equipped things, user devices, and servers
- data transport via access and backbone networks using advanced technologies
- data processing in real-time and/or use of storage data, either structured (databases) or unstructured (big data)
- data processing by using human-imitating technologies (machine learning, artificial intelligence)
- use of gained information in business processes to realize business value
These characteristics are displayed in the digital architecture model:
Digital architecture model
The digital architecture model aims to extend our understanding about digital value creation. It presents the relationships between the key concepts of digitization in the business context. The four structural components of a digital architecture - data sources, networks/cloud, artificial intelligence, business processes - and realized business value are described as follows.
Major categories of data sources are user devices, sensor-equipped objects, and servers for both private and business purposes. Digital devices provide artificial abilities to see, hear, locate, touch, feel, taste, smell, and move; they send significant data via networks for evaluation and application in business processes.
User devices are communication/computing equipment from private users or employees, such as smartphones, notebooks, desktop PCs, or workstations. Connected peripherals with sensors, e.g. smartwatches, can gather sensitive personal data (e.g. health status) generated by users. Very small, intelligent, embedded systems can be integrated into connected tools or other user objects (e.g. glasses).
Sensor-equipped endpoints collect data from natural and man-made objects (i.e. "Internet of Things"): living beings (people, animals, plants), vehicles (cars, bicycles, scooters), buildings (offices, factories, private house), household appliances and industrial machines (Industry 4.0).
Various types of data can continuously be collected from endpoints: audio (through microphones), video (through cameras), GPS data (geographical location, altitude, time, speed) and physical data through sensors (e.g. temperature, barometric pressure, humidity, brightness, chemical data of substances). In addition, tracking tools on user devices (e.g. cookies) help collect data about users and their behaviour.
The other category of data sources is servers. These can be public web servers connected to the internet or enterprise servers in secure environments and encrypted networks. Servers can be operated by private individuals, companies, government agencies, or other organizations. Data from public servers (e.g. social media, private websites, company pages, e-shops) can be retrieved by internet bots that simulate human interaction with the websites for analytics purposes. Web crawlers collect data from websites and index them to make their content accessible.
Networks and cloud storage
Networks are the transport systems for digital data. They link data sources and systems for data analyses (machine learning, artificial intelligence) and use (business applications). Networks move digital data streams between data sources, databases and application servers - independent of time and place.
Networks are divided into access and backbone networks. Access networks (e.g. 4G/5G Radio Access, LAN, WLAN) bridge data sources to long-distance backbone networks (the public internet or private networks, for example: MPLS WAN, Site-to-Site VPN). Networks also include short-distance wireless communication systems between user devices, e.g. RFID, NFC, Bluetooth, ultrasonic.
Networks allow digital data to be transmitted and processed in real time. If real-time data processing is not required or not possible, unstructured raw data are stored for later analysis in voluminous data stores, so-called data lakes (big data). Data storage can also be structured databases (e.g. data warehouse). The data are stored in company-managed data centers (private cloud) or at service providers (public cloud).
Artificial intelligence (AI) is the centerpiece of a digital architecture. AI gives meaning to the collected raw data. AI translates data from various sources into information that is valuable to the business.
AI strives to replace human cognitive abilities such as learning, thinking, and decision-making. The integration of AI with artificial sense organs (e.g. cameras or sensors) and human movement forms the basis of robots, i.e. simulated humans.
Machine learning (ML) is part of AI and is used to analyze patterns and relationships in data sets. The goal is to learn from experience and improve prediction accuracy over time through the use of models and algorithms. Data mining refers to ML learning methods for examining unstructured data sets, e.g. using regression analysis or clustering. Based on the ML results, alternative solutions can be suggested and/or decisions can be made that affect business activities.
The information obtained through AI can be used for primary or supporting activities of the generic value chain (Porter, 1985) to optimize various business functions (e.g. engineering, production, marketing, service). This business information may cause changes in product features, distribution, communication, pricing, etc. Business models and business architectures can be changed in whole or in part through the use of digital technologies. Examples of use cases are chat bots in customer communication, personalization in sales or automation in production. When changing business models and processes due to digitization, however, the focus must always be on value creation. The goal remains to increase customer benefits through products and services and thus improve profitability and the competitive position.
The effective use of digital data increases the value of the enterprise (e.g. decision making, process efficiency, supply chain optimization, understanding of customer needs and buying behavior, new marketing channels). Customer value results particularly from advanced product/service quality and easy purchase/delivery processing. Further, the effective integration of digital data into business processes leads to enhanced shopping experiences, e.g. tailor-made offers, simplified product searches, secure and fast payments, or additional services. Customers thus perceive economic, functional or psychological advantages.
Enterprises that neglect digitization today are likely to lose market share and risk disappearing from the markets entirely. On the other hand, companies that create new customer benefits with digital architectures can use it to expand their competitive advantage. An enterprise achieves a sustainable competitive advantage when its digital architecture is unique and difficult to imitate. In addition, the resulting services/products must generate more customer benefits than those of the competition.
Copyright 2022 Dr. C. Gellweiler - IT PM Consulting. All rights reserved.