Interdisciplinary integration of embedded systems as a core discipline
2026-04-06 02:07:58··#1
1. A Brief History of Embedded Systems 1.1 The Birth of Embedded Systems Embedded systems originated in the era of microcomputers. After a brief period of exploration in the embedded specialization of microcomputers, they entered the era of independent microcontroller development for embedded systems. Microcontrollers, intelligent electronic systems with processor cores, are developed directly based on embedded processors and peripheral integrated circuit technology. Even with a processor core, they are embedded processors, not general-purpose microprocessors. Figure 1 below illustrates that embedded systems are not dedicated computers. Figure 1: The Birth, Exploration, and Development of Embedded Systems Modern computers are microcomputers born on the basis of microprocessors. After the birth of microcomputers, their small size, low cost, and high reliability quickly led to the demand for intelligent control of large-scale electromechanical equipment. This required embedding microcomputers into large-scale electromechanical equipment to undertake the intelligent control of such equipment; thus, these microcomputers became dedicated computers embedded in specific electromechanical systems. To distinguish them from general-purpose computer systems, these dedicated computers were called "embedded computer systems." Based on the fact that the concept of embedded systems originated from microcomputers, early on, embedded systems could be considered dedicated computer systems. 1.2 The Failed Path of Dedicated Computer Exploration After the birth of embedded systems, in order to meet the most extensive embedded application requirements of target systems, the application models of embedded systems were constantly explored. Early development followed the approach of dedicated computers: industrial control computers, single-board computers, and microcomputers on a single chip. Industrial control computers were microcomputers that underwent mechanical and electrical hardening for embedded system applications, but they could not meet the basic requirements of embedded systems: small size, extremely low cost, high reliability, and good object coupling. Subsequently, board-level microcomputers (single-board computers) appeared, reducing computer size and price, and quickly sparking a wave of intelligent transformation of traditional electronic systems. Neither industrial control computers nor single-board computers could completely meet the requirements of small size, extremely low cost, and high reliability of embedded systems. Therefore, people directly simplified the microcomputer architecture and integrated it into a single semiconductor chip, creating a single-chip microcomputer. Motorola's 6801 series is a single-chip microcomputer integrated from a simplified version of the 6800 series microcomputer. Single-chip microcomputers completely solved the problems of extremely small size and extremely low cost of embedded systems, but there was no fundamental improvement in high reliability and object controllability. Internationally, chip-based microcomputers are called Single-Chip Microcomputers. For embedded applications in industrial control, high reliability and object coupling are crucial, exceeding the application requirements of traditional computers. For example, from the outset, embedded systems have considered "crashes" and "real-time performance" as important technical issues, and the bus, interface, and system configuration for electrical connection with the target system as important technological development directions. Therefore, embedded systems must break free from the constraints of "dedicated computers" and follow the independent path of "microcontrollers." Practice has proven that the development path of industrial control computers, single-board computers, and single-chip dedicated computers based on general-purpose microcontrollers is unworkable. 1.3 The Independent Development Path of Embedded Systems The development path of microcontrollers (MCUs) in embedded systems is a path of independent development, breaking free from the constraints of "dedicated computers." This is a path of independent development for microcontrollers pioneered by the Intel MCS51 microcontroller and the iDCX51 real-time multitasking operating system. The MCS51 is an original embedded processor developed based on microelectronics and integrated circuits, tailored to the requirements of embedded applications. The original architecture, control-oriented instruction set, Boolean space, external bus mode, and Special Function Register (SFR) management mode of the MCS51 laid the foundation for the hardware structure of embedded systems. The iDCX51 is an original real-time multitasking operating system specifically designed for the MCS51 microcontroller to meet the requirements of embedded applications. The MCS51 ushered in the era of independent microcontroller development for embedded systems. This was a microcontroller era, and foreign countries promptly and accurately changed "Single Chip Microcomputer" to "Microcontroller Unit." This was not a word game. The most important technological development hotspots in the era of microcontroller development for embedded systems were: expansion buses and communication buses that fully meet the requirements of embedded applications; real-time software requirements; sensor interfaces, servo drive interfaces, human-machine interfaces, and communication interfaces that connect with the target system; clock systems, power management systems, and low-power modes that meet low-power management requirements; and interrupt systems that can meet various state stimuli. In the microcontroller era, embedded systems were mainly used for the intelligent transformation of traditional electronic systems, forming an application era for semiconductor manufacturers and electronic engineers of target systems. More accurately, a microcontroller application system is an intelligent modern electronic system. Due to the widespread application of embedded systems and the various supports provided by semiconductor integrated circuits, DSP and PLD solutions have emerged in the field of embedded systems (see Figure 1). DSPs emphasize signal processing functions and, combined with embedded processors, have become an important branch of embedded systems; PLDs provide solutions for semi-custom gate array embedded application systems and have formed two major branches: SoPC and FPGA/CPLD. Regardless of the technological development direction, SoC is the unified category of embedded application systems. Currently, rejecting the concept of dedicated computers for embedded systems has significant practical implications. It is conducive to the correct positioning and cross-integration of the four pillar disciplines in embedded systems. 2 The Four Pillar Disciplines of Embedded Systems Currently, embedded systems have not yet formed an independent discipline system. From the birth of "embedded systems," the independent development path of microcontrollers, the connotation of microcontroller technology development, and the various solutions of embedded systems, "embedded systems" is the intersection and integration of four pillar disciplines, and is positioned and divided in a platform model. 2.1 Diagram of the Relationship between the Four Pillar Disciplines The four pillar disciplines of embedded systems are microelectronics, computer science, electronic technology, and object science. Their relationship is shown in Figure 2. Microelectronics is the foundation of embedded system development, the target discipline is the ultimate discipline for embedded system applications, and computer science and electronic technology are important guarantees for the development of embedded system technology. Figure 2 shows the intersection and integration of the four pillar disciplines of embedded systems. 2.2 Leading Microelectronics The leading role of microelectronics and semiconductor integrated circuits lies in providing the integrated circuit foundation for embedded system applications. Many important achievements of electronic technology and computer science are ultimately reflected in integrated circuits, from early digital circuit integration to today's mixed analog and digital, software/hardware combination, and IP-based knowledge and knowledge behavior integration. 2.3 Computer Science Serving the Platform After the emergence of modern computers, two major branches of computer science have formed: general-purpose computer science and embedded computer science. General-purpose computer science and embedded computer science have different technological development directions and technical connotations. Because embedded computer science is closely related to the target discipline and microelectronics, but the content of embedded computer science differs significantly from the original computer science, the concept of general-purpose computers cannot be used to interpret embedded systems. Therefore, embedded computer science should strengthen communication with microelectronics, electronic science, and the target discipline to jointly undertake the task of building the new discipline of embedded systems. In embedded systems, computer science undertakes the task of building embedded system application platforms, including integrated development environments, computer engineering methodologies, programming languages, and programming methodologies. 2.4 The Broadly Serving Electronic Technology Discipline: In embedded systems, electronic technology provides the broadest range of technical services. Electronic technology has rapidly advanced integrated circuit design in the microelectronics field from circuit integration, functional integration, and technical integration to knowledge integration; it provides hardware design technical support for embedded systems to computer science; and within the subject matter discipline, a large number of application engineers implement the widest range of applications on embedded hardware and software platforms. 2.5 The Ultimate Goal of the Subject Matter Discipline: The subject matter discipline is the end-user discipline of embedded systems. It encompasses almost all technological fields, forming an infinitely large application area for embedded systems. Today, embedded system experts no longer answer the question, "Where are embedded systems used?" For the subject matter discipline, an embedded system is merely an intelligent tool; the subject matter discipline aims to build an embedded application system within its field on top of an embedded system. The technical foundation of embedded application systems lies in the fundamental theories, application environment, and application requirements of this discipline. Simultaneously, in applications, technical requirements must be continuously raised for microelectronics, integrated circuit design, and embedded computer science disciplines in order to continuously improve the technical level of embedded system platforms. 3. Disciplines under the Platform Model The division of labor platform model is the basic development model of industry and technology in the knowledge economy era. All knowledge innovation and innovative knowledge application must follow the platform-based development path. 3.1 Origin of the Platform Model The platform model is a basic industrial and technological model in the knowledge economy era, and it is an inevitable phenomenon in the advanced stage of the development of the laws of separation and integration of human knowledge. It transforms the integrated industrial and technological model into a platform model under the medium of knowledge platforms. By comparing the radio industry in the 1960s with the VCD/DVD industry in the 1990s, we can find the essential difference between the integrated industrial model and the platform industrial model. In the 1960s, all companies in the radio industry, without exception, undertook the entire process from ideation, product development, prototype design, prototype production, to mass production, following a closed integrated industrial model. In the VCD/DVD era of the 1990s, a socialized division of labor emerged between semiconductor manufacturers and township enterprises: semiconductor manufacturers transformed VCD/DVD ideas into VCD/DVD hardware and software kits, while township enterprises purchased these kits and, with the technical support of semiconductor manufacturers, mass-produced VCD/DVD players. Semiconductor manufacturers never manufactured VCD/DVD players, and township enterprises never engaged in VCD/DVD technology research. The VCD/DVD hardware and software kits constituted the knowledge platform for VCD/DVD technology. Centered on this knowledge platform, a socialized division of labor between the knowledge industry and the manufacturing industry was achieved. The knowledge industry engaged in knowledge innovation, transforming innovative results into knowledge platforms, but did not engage in the final application of these results; the manufacturing industry, on the other hand, completed the final application of innovative results based on the knowledge platform. 3.2 The Platform Model of Embedded Systems Following the principle of the separation of knowledge development, knowledge innovators do not engage in knowledge application, and knowledge users do not need to understand the principles of innovative knowledge; following the principle of the integration of knowledge development, knowledge innovators should integrate innovative knowledge results into tools, transforming them into knowledge platforms, and knowledge users should realize the application of innovative knowledge based on these knowledge platforms. In early embedded systems, integrated circuit chips (microcontrollers and peripheral circuits) and development devices were application platforms provided by semiconductor manufacturers to users. Electronic engineers in the target field completed embedded system applications on these platforms. Currently, embedded systems, supported by four pillar disciplines, are inevitably forming an industrial and research ecosystem developed according to a platform-based division of labor. The target discipline is the end-user of embedded systems, and electronic application engineers in this field should implement embedded application system designs on an existing embedded system platform. Microelectronics, embedded computer science, and electronic technology (application engineers outside the target discipline) are not end-users of embedded systems. Their important task is to transform innovative scientific and technological achievements into various knowledge platforms. For example, integrated circuits in the microelectronics field provide advanced MCUs, peripheral chips, SoCs, and other IC platforms; embedded computer science provides integrated development environments, programming languages, algorithms, and computer engineering methodologies; and electronic engineers collaborate with microelectronics design, embedded computer science, and OEM manufacturers to complete the embedded system product platform and its technical services. From a platform perspective, an excellent embedded system product must be developed on an excellent embedded system platform. 3.3 Disciplinary Positioning and Division of Labor in the Platform Model Due to the interdisciplinary nature of the four pillar disciplines in embedded systems, each discipline demonstrates its strengths and weaknesses within embedded systems. Therefore, embedded systems present a problem of disciplinary positioning and cross-disciplinary integration. "Disciplinary positioning" reflects each discipline leveraging its own strengths to participate in the development of embedded systems based on its own disciplinary foundation. "Cross-disciplinary integration" involves continuously understanding the technical development requirements of other disciplines for embedded systems, based on disciplinary positioning, in order to build the best knowledge platform for embedded systems and achieve optimal embedded system applications. Due to the cross-disciplinary integration, the team building of each discipline should, on its own foundation, absorb a certain proportion of personnel from other disciplines. For example, in recent years, the field of integrated circuit design has absorbed many embedded application system design talents; when establishing embedded systems majors in computer science departments of universities, many application-oriented talents in embedded systems from relevant fields have been introduced. The positioning of the four pillar disciplines in embedded systems, in addition to the positioning of their disciplinary knowledge structure, should also reflect their positioning in the knowledge platform model. This positioning in the platform model is a 3+1 positioning. In other words, disciplines like microelectronics, computer science, and electronic technology build various application platforms for embedded applications without intervening in the specific applications of embedded systems. The target discipline, on the other hand, must realize the product application of embedded systems within its own field based on the embedded system application platform, without needing to intervene in the platform construction. For example, regarding embedded operating systems, the construction of the operating system should be positioned within the field of computer science. However, to build an excellent embedded operating system, one must understand the characteristics and environment of embedded applications and be able to foresee the firmware trends in future MCU chips. Within the target discipline, the operating system is viewed as a tool, requiring only an understanding of its performance and user interface. Microelectronics experts need to understand the characteristics of embedded operating systems and application software so that they can be incorporated into integrated circuit design once chip technology reaches a certain stage. Embedded systems represent a vast space, with immense potential for development in both embedded system platform construction and application. The key is to grasp one's own "positioning" and "division of labor," and to understand the "intersection" and "integration" of disciplines. References [1] He Limin. Industrial Model of Embedded Systems [J]. Microcontroller & Embedded System Application, 2006(1). [2] He Limin. The Modern Computer Industry Revolution from the Perspective of Embedded Systems [J]. Microcontroller & Embedded System Application, 2008(1). [3] He Limin. Multidisciplinary Integration and Penetration Centered on SoC [J]. Microcontroller & Embedded System Application, 2001(5).