Technical Articles
2025-01-13

Transformations from Embedded Systems to the Internet of Things

Transformations from Embedded Systems to the Internet of Things

After decades of development, embedded technology has been applied in various aspects of our lives. However, embedded systems have always had niche, highly specialized attributes, making it difficult for many from non-embedded fields to approach. In the past decade, the Internet of Things (IoT) has expanded into more fields, including home, business, industry, and agriculture, attracting not only those from the embedded field but also many from non-embedded areas into the IoT sector. From the development trends mentioned above, we can identify four significant transformations:

  • The technology involved in the IoT is becoming increasingly broad, raising the requirements for IoT development platforms.

  • The requirements for developers in the IoT have changed, placing greater emphasis on tools and usability.

  • IoT operating systems have transcended the traditional operating system kernel scope, beginning to integrate more capabilities.

  • The IoT is cloud-integrated, with rich application scenarios and business models, capable of aggregating more resources.

This article will focus on the four aspects of technology, people, operating systems, and business models, elaborating on the transition from embedded systems to the IoT and some reflections behind it.

Transformations from Embedded Systems to the Internet of Things

Transition from “Embedded Systems” to “Internet of Things”

1. Transformation of the Technology Stack
Embedded and IoT technologies have developed alongside microcontrollers, operating systems, cloud computing, artificial intelligence, and more. Based on the operational modes of devices, the development of technology can be divided into several stages:

  • Devices operate independently: This stage began in the 1970s, mainly involving technology fields such as microcontrollers, hardware development boards, peripheral drivers, and upper-level applications. It is primarily used in industrial control, such as monitoring and device indication. The most famous example is the 8-bit 51 microcontroller.

  • Multiple devices interconnect: With the development of wireless communication technologies like ad hoc, ZigBee (802.15.4), and low-power Bluetooth, embedded devices began forming networks. This stage is mainly used in smart grids and similar metering services. The technologies involved include not only those from the independent operation stage but also network connection technologies like 802.15.4 and low-power Bluetooth, as well as embedded operating systems like FreeRTOS, Contiki, and TinyOS.

  • Cloud integration stage: As the technological requirements for device control and management continuously increase, and with the development of cloud technology, more devices not only need to interconnect but also need to connect to the cloud. Wi-Fi modules and Wi-Fi and Bluetooth combo modules have emerged. The author believes that it is from this stage that we truly enter the IoT phase. Through gateway technologies and protocols like MQTT/CoAP, devices can connect to the cloud and be managed via mobile phones. Compared to the previous stage, more diverse connection technologies such as MQTT/CoAP and cloud technologies like IoT cloud platforms have been introduced.

  • IoT smart devices: With technological advancements, the devices are no longer limited to Wi-Fi and Bluetooth connection types; smart speakers, which were particularly popular a few years ago, and IP cameras with hundreds of millions of shipments annually have emerged. These devices are characterized by multimedia requirements and the ability to consume more cloud resources, including not only storage resources but also computation and artificial intelligence (AI) algorithm resources. Compared to the previous stage, multimedia technology, streaming network technology, and cloud AI technology have been added, raising the breadth and depth of technical requirements for developers building IoT capabilities.

From the above analysis, we can see that the content of the technology stack is becoming richer and the requirements are increasing. This raises the question: do IoT developers need to understand all these technologies to innovate and develop? The author’s answer is certainly not; below, the author will elaborate on their understanding of the requirements for IoT developers.
2. Transformation of Developers
The author categorizes IoT developers into two types: those building capabilities for IoT platforms and those developing based on the capabilities provided by IoT platforms.
The first type of developer needs to understand the technologies required across the entire IoT field. As IoT technology evolves, the requirements for this type of developer have become very high. They need to grasp the full chain of technology from the device side to the cloud side and find their place within it. If they do not systematically understand IoT technology from a cloud-integrated perspective but rather follow traditional embedded thinking, they will gradually be eliminated as technology continues to evolve. A key goal for these developers in building IoT platforms is to attract more second-type developers to join and enrich the IoT ecosystem.
Compared to the first type of developer, the second type also faces significant challenges. This group mainly comes from two sources: traditional embedded developers and internet developers. Both groups need to learn more knowledge to better develop based on IoT platforms. Traditional embedded developers need to understand more multimedia, cloud, and artificial intelligence technologies. Internet developers need to comprehend the various limitations on embedded devices to better engineer cloud, multimedia, and AI technologies into IoT scenarios.
The author believes that the mission of the first type of developer is to help more second-type developers enter the IoT field. The key is for the first type of developer to achieve the following two points:

  • Rich functional components: Including various hardware modules, device-side and cloud-side software functional modules, and accompanying development and debugging tools.

  • Low-code development: The full utilization of scripting languages like Python and JavaScript in the IoT field can significantly lower the barriers for developers.

As IoT technology continues to mature and foundational platforms and development tools improve, the author believes that the first type of developer will become increasingly rare, while the second type will grow in number. This will make IoT development no longer just a domain for a small number of specialized developers; students, front-end developers, product managers, and technology enthusiasts will all be contributors to the IoT.

Transformations from Embedded Systems to the Internet of Things

3. Transformation of Operating Systems
Since operating systems are a crucial foundational technology in the IoT field, and currently, IoT operating systems are diverse, this section will specifically discuss the operating systems. Based on the evolving stages of embedded and IoT technologies, the author categorizes operating systems into four stages:

  • No operating system: Early embedded devices, due to their simple logic, did not require an operating system.

  • Simple operating systems: As we moved into the multiple device interconnect stage, operating systems specifically designed for IoT devices emerged, such as Contiki and TinyOS. These operating systems are designed for data and event reporting from devices, characterized by their simplicity and event-driven design. They can respond quickly to events that need reporting, with programming logic designed accordingly.

  • Real-time operating systems: With the continuous development of the embedded field, real-time embedded operating systems have emerged, with FreeRTOS being the most typical representative. These operating systems are known for their better real-time performance compared to Linux and more comprehensive kernel functionalities, but they do not have a strong ecosystem and limited upper application support.

  • IoT operating systems: With the advent of the IoT, real-time operating systems primarily provide kernel capabilities, while network, multimedia, configuration tools, and application ecosystems struggle to meet IoT demands. This has led to the emergence of IoT operating systems such as RT Thread, LiteOS, AliOS Things, as well as Linux and Android. The author believes that for some time, many operating systems will coexist and it will be difficult to unify them. The main reason is that chip manufacturers currently have no motivation to pursue unification; each can choose the one that best suits their needs, and switching to another operating system brings limited advantages.

Some viewpoints suggest that to solve the fragmentation issue of the IoT, we should start with a unified operating system. The author disagrees with this viewpoint. The IoT is inherently fragmented; this is both a challenge and part of its charm. To illustrate this point, consider an extreme example: if unifying the operating system could resolve the IoT fragmentation issue, we would directly unify it at the chip level. However, this is clearly impossible. The greatest value of developing IoT operating systems lies in achieving extreme performance and cost advantages through deep integration of software and hardware. Since there are already diverse IoT operating systems, let them exist. The solution to the IoT fragmentation issue does not lie in the operating system.
4. Transformation of Business Models
In the traditional embedded field, companies primarily made profits by selling hardware, which made it challenging to support a publicly listed company, let alone one with a market value of billions or even hundreds of billions. However, in the IoT field, due to its integrated end-to-cloud characteristics, the entire business model has undergone a significant transformation from one-time hardware sales in the embedded field to operational profit models, such as selling cloud storage, cloud services, and content. Besides the operational revenue brought by IoT itself, it also serves as a strong moat for the mobile ecosystem, further increasing the added value of mobile devices as super terminals. It is believed that more billion-dollar IoT companies, even those valued at hundreds of billions, will emerge, as those with such market values will have effectively addressed the IoT fragmentation issue and formed a strong ecosystem, achieving explosive growth.

Transformations from Embedded Systems to the Internet of Things

5. Conclusion
As the evolution from embedded systems to the IoT progresses, the complexity of technology increases. However, for IoT developers, the barriers will inevitably lower, allowing more developers to create innovative products based on simple scripting languages, rich ecological libraries, and tools to meet demands for convenient living and efficient production. Furthermore, the development of the IoT is not just a technical evolution but a transformation of business models. Through technological innovation, the barriers for developers have been lowered, making IoT development no longer a niche and geeky domain but an accessible innovation base for the general public. This shift, combined with the rapid advancements in cloud technologies, has transformed the IoT business model from relying primarily on selling development boards to selling services, cloud resources, and innovative products. This mass and high-value business model supports the emergence of billion-dollar and even hundred-billion-dollar companies in the IoT space.

Get your free eBook

“Navigating RocketMQ in the Spring Ecosystem”

How can you navigate RocketMQ within the Spring ecosystem? What are the various approaches? This book begins with the story of Luo Meiqi (RocketMQ) and Chun Bote (SpringBoot), detailing the three mainstream methods of using RocketMQ in the Spring ecosystem, guiding you through practical operations!

Scan to add Ali Mei as a friend and reply “Navigate” to get it~ (If the scan is invalid, you can add via WeChat IDs alimei4, alimei5, alimei6, alimei7 directly)
Transformations from Embedded Systems to the Internet of Things

Related posts

Leave a Comment

Your email address will not be published. Required fields are marked *