Fluent programming languages used to write operating systems, such as C: understanding of component-level hardware; and the ability to easily cross software and hardware, these are all core skills of today's embedded engineers.
Embedded development includes the diversity of technologies and the ever-expanding functions, so what kind of skills are important for engineers to be competitive, making them competitive in the job market and developing their careers.
No need to be an electrical engineer or chip designer, but you need to understand some hardware. Understand the working principles of address and data buses, interface buses such as I2C and SPI, and GPIO. You should also have a basic electronics background on GPIO and its related voltages. This can help you understand the basics of connecting embedded hardware to the real world.
In addition, you also need to determine whether you are interested in device drivers, porting the operating system, or letting the operating system run from scratch. These all require the ability to read the manufacturer's data sheet, know what a register is and how to manipulate the bits in the register. On the other hand, if you only want to focus on software, is the application only embedded (ie, resource limited) or has real-time and/or real-time capabilities? Impact critical to safety? Embedded, real-time and security are essential. They are relevant in many ways, but they are not the same thing. They all require slightly different skills. All these skills can be learned. However, most people cannot try to solve all these areas, especially in the early stages of their careers. However, a plan can be developed that starts with "justice" embedded in resource constraints. Then, go to real-time (soft real-time and hard real-time). Finally, it is essential to safety (the most demanding real-time domain due to poor design).
Today's embedded engineers are expected to be familiar with open source technologies such as Linux, Zephyr, Tensorflow or RISC-V. In fact, open source technologies and licensing are becoming ubiquitous in embedded software. In addition, SoC-level software optimization skills are increasingly important for performance, memory and power optimization. As embedded engineers are required to use dedicated hardware accelerators (such as GPU, DSP, video, vision, security, and neural network accelerators) to do more, system-level thinking will become more important.