Thirteen stages, seven goals, one honest path

What you study

• Control flow, functions, recursion and modular thinking
• Data types, scope, arrays and strings
• Problem decomposition and pseudocode
• Version control basics with Git

Why it matters

Embedded work is still software work. Weak fundamentals show up later as "firmware bugs" that are really C mistakes.

Prerequisites

• Basic comfort with a computer and command line

Practical exercise

Solve 30 small problems and refactor one solution three times, documenting each improvement.

Skills gained

Recommended next step

Move to Embedded C once you can write small programs without looking up syntax.

Effort indicator: 3–5 weeks at 6–8 hours/week

What you study

• Pointers, pointer arithmetic and function pointers
• Memory layout: stack, heap, BSS, data, text
• const, volatile and compiler optimisation effects
• Bit manipulation, masks and register-style access
• Structures, unions, bit-fields and the preprocessor
• Compilation, linking and reading map files

Why it matters

This is the language interviewers test hardest and the language your firmware will actually be written in.

Prerequisites

• Stage 1 — programming foundations

Practical exercise

Implement a ring buffer and a CRC routine, then explain the memory layout of both from the map file.

Skills gained

Recommended next step

Apply C to real silicon in Microcontroller Fundamentals.

Effort indicator: 5–7 weeks at 6–8 hours/week

What you study

• Processor core, bus matrix, memory map and clock tree
• GPIO, timers, interrupts, ADC, PWM and DMA
• Datasheet and reference-manual reading
• Power modes and low-power considerations

Why it matters

Register-level understanding is what separates an embedded engineer from someone who calls library functions.

Prerequisites

• Stage 2 — embedded C
• Basic electronics

Practical exercise

Bring up an unseen board from documentation alone: clocks, GPIO, then a timer interrupt.

Skills gained

Recommended next step

Start talking to external chips in Peripheral Interfaces.

Effort indicator: 4–6 weeks at 6–8 hours/week

What you study

• UART framing, flow control and DMA reception
• SPI modes, chip selects and throughput
• I2C addressing, clock stretching and error recovery
• Logic analyser workflow and waveform interpretation
• Sensor datasheets and timing diagrams

Why it matters

Most field failures are on a wire, not in an algorithm. Interface debugging is a daily skill.

Prerequisites

• Stage 3 — microcontroller fundamentals

Practical exercise

Interface two sensors plus a display, capture every bus transaction on a logic analyser, and document error-recovery behaviour.

Skills gained

Recommended next step

Structure the code into Firmware Development.

Effort indicator: 4–5 weeks at 6–8 hours/week

What you study

• Layered firmware architecture and driver interfaces
• Interrupt-driven design and non-blocking patterns
• State machines and event-driven structure
• Startup code and linker script fundamentals
• Unit testing and host-side test harnesses

Why it matters

Projects stop working at scale unless the code has structure. This is where hobby code becomes engineering.

Prerequisites

• Stage 4 — peripheral interfaces

Practical exercise

Refactor a flat superloop application into layered drivers plus a documented state machine with unit tests.

Skills gained

Recommended next step

Deepen on a real target in ARM-based Systems.

Effort indicator: 5–7 weeks at 6–8 hours/week

What you study

• ARM Cortex-M architecture, exceptions and faults
• STM32 ecosystem and register-level peripheral drivers
• ESP32 architecture and dual-core considerations
• Toolchains, flashing, OpenOCD/debug probe workflows
• Hard-fault analysis and low-level debugging

Why it matters

ARM Cortex-M dominates the industry. Depth here transfers directly to most job descriptions.

Prerequisites

• Stage 5 — firmware development

Practical exercise

Write a custom linker script, trigger and analyse a hard-fault, then fix the root cause.

Skills gained

Recommended next step

Add concurrency with RTOS.

Effort indicator: 5–7 weeks at 6–8 hours/week

What you study

• Scheduling, context switching and priorities
• Queues, mutexes, semaphores, event groups
• Interrupt-safe APIs and deferred processing
• Priority inversion, deadlock, starvation, stack sizing
• Watchdogs, CPU-load and timing analysis

Why it matters

Concurrency is the single most common source of interview failure and field defects in embedded roles.

Prerequisites

• Stage 6 — ARM-based systems

Practical exercise

Build a multi-task controller with queue-based IPC, then deliberately inject a priority inversion and fix it.

Skills gained

Recommended next step

Choose a specialisation: Embedded Linux, IoT, automotive or Edge AI.

Effort indicator: 6–8 weeks at 6–8 hours/week

What you study

• Linux fundamentals, user/kernel space and syscalls
• Cross-compilation and toolchains
• Boot process, bootloaders and kernel command line
• Kernel architecture, modules and device trees
• Character device drivers and driver debugging
• Build systems: Buildroot and Yocto

Why it matters

Gateways, cameras, infotainment and edge gateways all run Linux. Driver skill is a scarce, well-paid specialty.

Prerequisites

• Stage 5 — firmware development
• Linux command-line comfort

Practical exercise

Implement a character driver exposing a device, add its device-tree node, and boot an image you built.

Skills gained

Recommended next step

Connect devices in IoT and Networking.

Effort indicator: 8–11 weeks at 6–8 hours/week

What you study

• TCP/IP model, sockets and connection handling
• MQTT topics, QoS and session behaviour; CoAP alternatives
• Device provisioning, identity and onboarding
• Telemetry design, batching and back-off
• OTA firmware updates and rollback safety
• Security fundamentals for connected devices

Why it matters

Connected products fail at the network edge. Knowing retries, offline behaviour and updates is real engineering.

Prerequisites

• Stage 4 — peripheral interfaces
• Stage 5 — firmware development

Practical exercise

Ship telemetry from a device to a broker, survive a network outage, and perform a validated firmware update.

Skills gained

Recommended next step

Combine with AI in Embedded AI, or go deeper in Automotive.

Effort indicator: 5–7 weeks at 6–8 hours/week

What you study

• ECU categories, network topology and gateways
• CAN, CAN FD, LIN and automotive Ethernet awareness
• UDS diagnostics and diagnostic message analysis
• AUTOSAR layering concepts
• Functional safety (ISO 26262) and cybersecurity awareness

Why it matters

Pune and Maharashtra host large automotive engineering clusters; diagnostics and CAN are recurring hiring requirements.

Prerequisites

• Stage 7 — RTOS
• Stage 9 basics of CAN from peripheral interfaces

Practical exercise

Build a CAN data logger that performs a UDS session to configure its own capture settings.

Skills gained

Recommended next step

Apply AI at the edge for perception or condition monitoring.

Effort indicator: 5–7 weeks at 6–8 hours/week

What you study

• ML fundamentals and problem framing for sensors
• Data collection, windowing, labelling and preprocessing
• Model selection and evaluation without data leakage
• Quantisation, compression and memory budgeting
• TinyML runtimes and on-device inference
• Latency, power and system-level testing

Why it matters

Edge AI only counts when it runs inside real constraints. This stage is about measurements, not demos.

Prerequisites

• Stage 2 — embedded C
• Stage 5 — firmware development
• Comfort with Python for training

Practical exercise

Train, quantise and deploy a classifier, then report on-device latency, RAM and accuracy side by side.

Skills gained

Recommended next step

Integrate everything in Integrated Projects.

Effort indicator: 7–9 weeks at 6–8 hours/week

What you study

• Requirements capture and constraint budgets
• Architecture across MCU, RTOS/network and model layers
• Integration testing and hardware-in-the-loop validation
• Measurement: timing, memory, power, reliability
• Documentation and demonstration

Why it matters

Interviewers probe integration experience, not completed tutorials. This is where evidence is produced.

Prerequisites

• Any 7+ preceding stages depending on the chosen pathway

Practical exercise

Deliver one end-to-end system with a written design, measured results and a recorded demonstration.

Skills gained

Recommended next step

Convert the work into applications in Career Preparation.

Effort indicator: 6–10 weeks per substantial project

What you study

• Technical resume and project write-ups
• GitHub portfolio structure and README quality
• Timed Embedded C and debugging assessments
• RTOS and Embedded Linux interview preparation
• Project explanation practice and mock interviews
• Job-search strategy and tracking

Why it matters

Skill without evidence is invisible to recruiters. Evidence without interview skill is wasted.

Prerequisites

• At least two completed integrated projects

Practical exercise

Run three mock interviews and publish two project write-ups that explain trade-offs honestly.

Skills gained

Recommended next step

Continue targeted study based on the roles you are actually targeting.

Effort indicator: 4–6 weeks, running alongside applications