From Embedded to IoT: Evolving Tech Stacks, Developer Roles, OS & Business Models

After decades of development, embedded technology has permeated many aspects of daily life, but it has remained a niche with strong specialization. In the past ten years, the Internet of Things (IoT) has spread across home, commercial, industrial, and agricultural fields, attracting both traditional embedded engineers and newcomers from other domains. Four major shifts can be observed.

Technology Stack Shift

The evolution of embedded and IoT technologies follows the development of microcontrollers, operating systems, cloud computing, and AI. The device operation modes can be divided into four stages:

Device standalone: originating in the 1970s, focusing on microcontrollers, development boards, peripheral drivers, and simple applications such as industrial control and monitoring.

Multiple devices interconnect: with the rise of wireless technologies like ad‑hoc, ZigBee, and low‑power Bluetooth, embedded devices form networks for smart grid and meter‑reading scenarios, using protocols and OSes such as FreeRTOS, Contiki, TinyOS.

Cloud‑integrated stage: devices require not only networking but also cloud connectivity via Wi‑Fi, BT‑combo modules, gateways, MQTT/CoAP, enabling remote management and richer services.

Smart IoT devices: modern devices (e.g., smart speakers, IP cameras) demand multimedia capabilities, cloud storage, compute, and AI algorithms, expanding the required skill set for developers.

This increasing richness and complexity raise the question of whether every IoT developer must master all these technologies. The answer is no, as explained in the following sections.

Developer Shift

IoT developers can be split into two groups: platform builders who provide core IoT capabilities, and application developers who build solutions on top of those platforms.

Platform builders need a comprehensive understanding of the entire IoT stack—from device to cloud—to create ecosystems that attract the second group. Their mission includes offering rich functional components (hardware modules, device‑side and cloud‑side software, debugging tools) and enabling low‑code development with languages such as Python and JavaScript, thereby lowering entry barriers.

Application developers come from two backgrounds: traditional embedded engineers who must learn multimedia, cloud, and AI technologies, and web/Internet developers who must grasp device constraints and engineer cloud‑centric, multimedia‑rich solutions.

Operating System Shift

Operating systems are a foundational technology in IoT, and their evolution can be categorized into four phases:

No OS: early embedded devices with simple logic did not require an operating system.

Simple OS: event‑driven systems like Contiki and TinyOS designed for data reporting and low‑power operation.

Real‑time OS: kernels such as FreeRTOS offering better real‑time performance but limited ecosystem support.

IoT OS: platforms like RT‑Thread, LiteOS, AliOS Things, as well as Linux and Android, providing networking, multimedia, provisioning tools, and application ecosystems. Fragmentation remains inevitable because chip vendors choose the OS that best fits their hardware. Some argue that unifying the OS would solve IoT fragmentation, but the author disagrees, emphasizing that fragmentation is intrinsic to IoT’s diversity and that the value of an IoT OS lies in deep hardware‑software integration for performance and cost advantages.

Business Model Shift

Traditional embedded markets relied on hardware sales, limiting company scale. IoT’s cloud‑integrated nature transforms the model into recurring revenue streams: cloud storage, services, and content. This shift enables companies to achieve billion‑dollar valuations by monetizing ongoing services rather than one‑time hardware transactions.

Conclusion

As embedded systems evolve into IoT, technical complexity rises while development barriers fall thanks to scripting languages, rich ecosystems, and powerful tools. The transition also reshapes business models from hardware‑centric sales to cloud‑based services, opening opportunities for a broader range of developers—including front‑end engineers, product managers, and hobbyists—to innovate and create value in the IoT era.