Shivansh

EEE Student | Embedded System & Firmware

Built ELECTROSENSE — ESP32, Edge AI & Capacitor Health Monitoring

STM32 • CAN Bus • RTOS
🔗

[Shivansh @firmware_core]

FreeRTOS Mutex Implementation on STM32

Implemented & tested a Mutex-based synchronization project using 3 LEDs.

Learned:
- Protecting shared resources
- Synchronizing multiple tasks
- Managing task access

•Practice with FreeRTOS on STM32

#STM32 #FreeRTOS #EmbeddedSystems

[Shivansh @firmware_core]

@TechCricforlife @pohawithketchup 🤣🤣

[Shivansh @firmware_core]

My STM32 learning Journey

• GPIO: LEDs & buttons
• UART: TX/RX & serial commands
• ADC: analog i/p & DMA
• I2C: SSD1306 OLED communication

•Learning embedded firmware with STM32CubeIDE & HAL.

Next: SPI, CAN & RTOS.

#STM32 #EmbeddedSystems
#firmware

[Shivansh @firmware_core]

Built a custom 5x7 font engine for an SSD1306 OLED on STM32 Nucleo—completely from scratch! 🛠️️
No libraries, just raw HAL I2C, custom glyph bytes, and pixel-level rendering from a plain C string.

Who else loves ditching standard dependencies?

#STM32 #EmbeddedSystem
#Firmware

[Shivansh @firmware_core]

Result: 95.2% accuracy, real-time, fully on-device — no cloud dependency.
Repository (firmware, dataset, dashboard, demo video):
https://t.co/WV3MkfGFyA

[Shivansh @firmware_core]

Built a custom 5x7 font engine for an SSD1306 OLED on STM32 Nucleo—completely from scratch! 🛠️️
No libraries, just raw HAL I2C, custom glyph bytes, and pixel-level rendering from a plain C string.

Who else loves ditching standard dependencies?

#STM32 #EmbeddedSystem
#Firmware

[Shivansh @firmware_core]

@TechCricforlife It's done bro 🙂

[Shivansh @firmware_core]

@TechCricforlife 🤣🤣🤣

[Shivansh @firmware_core]

@TechCricforlife Ghee refill day 3 🫢

[Shivansh @firmware_core]

@TechCricforlife @TechCricforlife
Bhai teri gf bhi thi 🥺

[Shivansh @firmware_core]

Result: 95.2% accuracy, real-time, fully on-device — no cloud dependency.
Repository (firmware, dataset, dashboard, demo video):
https://t.co/WV3MkfGFyA

[Shivansh @firmware_core]

Built an ESP32-based system for predictive capacitor health monitoring — detecting degradation before failure occurs, not after.
95.2% classification accuracy across three states (healthy / faulty / no_capacitor), running fully on-device.
Technical breakdown below:

[Shivansh @firmware_core]

FreeRTOS was necessary rather than optional here: the ML model was trained on uniform 30Hz samples. Any blocking operation in a single-loop design (e.g., display refresh) would introduce sampling jitter, causing inference-time feature distributions to diverge from training data.

[Shivansh @firmware_core]

@TechCricforlife @TechCricforlife 🤮

[Shivansh @firmware_core]

@TechCricforlife 😂

[Shivansh @firmware_core]

@gentleman07_ @TechCricforlife

[Shivansh @firmware_core]

@TechCricforlife 🤣🤣