🛰️ VELA Flight Software

A fault-tolerant, RTOS-based embedded flight software architecture designed to model modern spacecraft data systems. VELA bridges bare-metal C programming on ARM Cortex-M4 processors with a high-level Python Ground Data System (GDS), featuring hardware-level fault isolation, cryptographic command authentication, and Edge AI anomaly detection.

Designed to demonstrate production-ready aerospace engineering principles for startups, defense labs, and autonomous systems research.

---

📊 Visual Gallery

Hardware Implementation	IDE Configuration
STM32 Black Pill (Flight PC) and ESP32 (Payload) connected to the host.	Bare-metal Cortex-M4 pinout mapping in STM32CubeIDE.

Python Ground Data System	Nominal Mission Dashboard
Python GDS actively unpacking 17-byte binary telemetry frames.	Real-time thermal and attitude telemetry streaming via Streamlit.

AI Anomaly Detection Triggered
TinyML Edge AI model successfully catching a simulated 500°C catastrophic sensor failure and triggering an automated alert.

---

⚙️ System Architecture & Data Flow

The system operates across two independent microcontrollers and a host machine, connected via UART and synchronized through a preemptive real-time operating system.

graph TD;
    subgraph Payload Node
        E[ESP32 Sensor Sim] -->|17-Byte Binary Packets| UART1
    end
    
    subgraph Flight Computer STM32F401
        UART1[UART ISR] --> C[Comms Task]
        C -->|Thread-Safe Queue| T[Telemetry Task]
        C -->|Thread-Safe Queue| AI[TinyML AI Task]
        W[FDIR Watchdog & MPU] -.- C
        W -.- T
    end
    
    subgraph Ground Data System
        T -->|Serial COM| G[Python Data Ingestion]
        G -->|Auto-Logging| CSV[(Telemetry Dataset)]
        CSV --> D[Streamlit Web Dashboard]
    end

Loading

---

✦ Core Subsystems & Capabilities

• Preemptive Multitasking (FreeRTOS): Concurrent task execution managing communications, telemetry processing, and background AI inference without blocking.
• FDIR & Hardware Fault Isolation: An independent watchdog monitors thread starvation. The ARM Memory Protection Unit (MPU) locks critical SRAM zones, intercepting rogue pointers to trigger a hardware-level MemManage exception and an automated safe-mode reboot.
• Space Cybersecurity (SpaceSec): Implementation of a lightweight FNV-1a Hash-MAC cryptographic engine. All incoming uplink commands are verified against a pre-shared secret key to prevent RF command spoofing.
• Edge AI (TinyML): A bare-metal statistical anomaly detection model trained on historical telemetry. The AI continuously calculates Z-scores on live data streams, triggering hardware visual alarms if sensors deviate beyond a 3-sigma operational envelope.
• Automated Data Dictionary: NASA JPL-style YAML-to-C code generation (dict_generator.py) ensures zero mismatch between the flight computer's C-structs and the Python unpacking logic.
• Live Mission Dashboard: A real-time Streamlit web interface that ingests serial CSV logs to visualize thermal walks, radiation metrics, and orbital attitude dynamics.

---

🔌 Hardware Requirements

To replicate or deploy this project, the following hardware is required:

• Flight Computer: STM32F401CCU6 "Black Pill" (ARM Cortex-M4)
• Payload Simulator: ESP32 Development Board
• Interface: CP2102/FTDI USB to TTL UART Bridge
• Wiring: Standard Dupont jumper cables (Common Ground, TX/RX crossover)

---

💻 Installation & Quick Start

1. Software Prerequisites

• STM32CubeIDE (For compiling and flashing the STM32)
• Arduino IDE or PlatformIO (For flashing the ESP32)
• Python 3.10+

2. Clone the Repository

git clone [https://github.com/yourusername/vela-flight.git](https://github.com/yourusername/vela-flight.git)
cd vela-flight

3. Flash the Hardware

1. Open sim_node/ in Arduino IDE, select your ESP32 board, and upload.
2. Open the flight_software/ folder as a workspace in STM32CubeIDE.
3. Build the project (Release or Debug) and flash it to the STM32 Black Pill via ST-Link or DFU.

4. Boot the Ground Data System (GDS)

Install the required Python data and UI libraries:

python -m pip install pyserial PyYAML pandas streamlit

Terminal 1: Start Telemetry Ingestion

python tools/ground_station.py
# Select the COM port connected to your STM32/ESP32 bridge.

Terminal 2: Launch the Web Dashboard

python -m streamlit run tools/web_dashboard.py

The dashboard will automatically open in your default web browser at http://localhost:8501.

Made by aman :)