An embedded system is a computer hiding inside a product that does not look like a computer. Your microwave, the ABS module in your car, a pacemaker, an ESP32 in a smart bulb, the SoC inside a Wi-Fi router, the flight controller of a drone, the firmware on a credit-card chip. Each of them is a complete machine: processor, memory, peripherals, code, often a battery, sometimes a radio. None of them advertise themselves as computers. The owner expects them to "just work" for years, often decades, with no input.
This chapter is about what it takes to make those things actually work. Chapter 14 covered general computer architecture (datapath, pipeline, cache). Chapter 15 met specific microcontrollers (8051, AVR, ARM Cortex-M, MSP430). Here we zoom out to the system, the things sitting around the chip and the firmware running on top: power, sensors, serial buses, RTOS scheduling, bootloaders, OTA updates, IoT protocols, edge AI, and the security holes that all of those open up. Chapter 24 (hardware security) will exploit many of the surfaces we are about to install.
The mental shift you need is this: a PC is built to run anything you throw at it, and to crash gracefully when it does not. An embedded system is built to do one thing, run forever, and never crash because there is no human standing next to it to press a reset button. That single shift in expectation drives every design decision below.