Design and implement a distributed embedded system for real-time vehicle telemetry, capable of operating in both connected and offline environments, with local buffering and cloud-based visualization.
Context
This project simulates an automotive telemetry ECU for fleet monitoring.
The system continuously collects vehicle data (temperature, RPM, speed, battery), stores it locally during network outages, and synchronizes it with a cloud pipeline when connectivity is available.
It demonstrates:
- real-time embedded data acquisition
- distributed system design (multi-MCU architecture)
- fault-tolerant data handling
- IoT cloud integration and visualization
System Overview
The architecture is based on a distributed multi-microcontroller system:
- STM32 → Real-time acquisition & processing
- ESP32 → Connectivity, storage, and cloud interface
- Server → Data pipeline & visualization
System Architecture
Global system architecture: STM32 (data acquisition) → ESP32 (gateway) → MQTT → Node-RED → InfluxDB → Grafana
Key Features
- Real-time sensor acquisition (temperature, speed, RPM, battery simulation)
- Multi-MCU communication over UART (compact JSON frames)
- Offline data buffering using circular storage (LittleFS)
- Automatic data synchronization when network is restored
- End-to-end IoT data pipeline (MQTT → Node-RED → InfluxDB → Grafana)
- Real-time dashboard visualization with alert thresholds
- Embedded watchdog for system reliability (planned)
Embedded Architecture Breakdown
STM32 - Real-Time Acquisition Node
- reeRTOS-based firmware
- Acquisition task (100 Hz)
- Watchdog task
- Sensor acquisition via ADC / I2C (BME280, inputs simulation)
- HAL drivers (ADC, GPIO, UART, I2C)
- Compact JSON serialization of data
- UART transmission to ESP32
ESP32 - IoT Gateway & Storage Node
- FreeRTOS-based communication architecture
- UART reception task
- MQTT publishing task
- Storage management task
- MQTT client (Mosquitto broker)
- Local storage using LittleFS (circular buffer for offline mode)
- Automatic reconnect & data sync
- OTA support (planned)
Cloud & Visualization Layer
- MQTT Broker: Mosquitto
- Data Processing: Node-RED
- Time-Series DB: InfluxDB
- Visualization: Grafana
Hardware
- STM32F103RCT6 (STM32Cube framework)
- ST‑Link V2
- ESP32 WROOM 32 (Arduino framework)
- BME280 (I2C)
- Joystick + Potentiometer + 2 Buttons
- 18650 battery + holder
- Breadboard + resistors + LEDs
Software implementation
STM32 Firmware
ESP32 Firmware
Visualization & Dashboard
Demo
Repositories
STM32 Firmware (access grant on request): Repository
ESP32 Firmware (access grant on request): Repository
Next Steps
- Improve data compression and optimization
- OTA update
- Extend sensor integration (CAN / OBD real interface)
- Add security (TLS MQTT, authentication)
- Watchdog for system reboot
Technical Highlights
- Distributed embedded system (multi-MCU architecture)
- Real-time firmware (FreeRTOS, 100 Hz acquisition)
- Fault-tolerant storage (offline buffering + sync)
- End-to-end IoT pipeline
- Embedded-to-cloud integration
- Modular firmware architecture
