CORAL is an industrial-grade space safety platform engineered to process Conjunction Data Messages (CDMs), perform non-linear uncertainty propagation, and manage close-approach warning logs for active satellite missions.
The platform combines a high-speed orbital mechanics engine written in C++17 Differential Algebra (DA) with a modern FastAPI REST Server and structured SQLite persistence.
-
High-Accuracy Propagation: Integrates an adaptive step-size Runge-Kutta 7(8) Dormand-Prince (
RK78) solver with a PI step controller. - Non-Linear Covariance Mapping: Uses Order-1 (Jacobian) and Order-2 (Hessian / State Transition Tensors) Differential Algebra to map complex non-linear uncertainty structures over time.
- Robust State Ingestion: Features an Extended Kalman Filter (EKF) update implemented in the Joseph Stabilized Form to prevent matrix asymmetry due to numerical rounding.
-
Multi-Method Risk Valuation: Computes automated collision probabilities (
$P_c$ ) via Alfano, Foster, Chan, and advanced Monte Carlo models. -
Interactive Visualization Engine: Generates interactive Plotly charts tracking
$P_c$ time-series curves, risk profiles, and 2D target altitude comparisons. -
Mission Maneuver Advice: Automatically checks safety thresholds to provide active
$\Delta V$ magnitude, direction, and justification recommendations.
┌─────────────────────────────────────┐
│ USER INTERFACES │
│ ┌─────────────┐ ┌─────────────┐ │
│ │ Streamlit │ │ FastAPI │ │
│ │ Dashboard │ │ REST API │ │
│ │ (Port 8501)│ │ (Port 8000)│ │
│ └──────┬──────┘ └──────┬──────┘ │
│ │ │ │
│ └───────┬────────┘ │
│ │ │
└─────────────────┼──────────────────┘
│
┌─────────────────▼──────────────────┐
│ DATA & ANALYSIS LAYER │
│ ┌─────────────┐ ┌─────────────┐ │
│ │ SQLite │◄───│ CDM_parser │ │
│ │ (WAL mode) │ │ (.py) │ │
│ └──────┬──────┘ └─────────────┘ │
│ │ │
│ ┌──────┴──────┐ ┌─────────────┐ │
│ │ Collision_ │ │ Risk_Trend │ │
│ │ Probability│ │ (.py) │ │
│ │ (.py) │ │ │ │
│ └──────┬──────┘ └──────┬──────┘ │
│ │ │ │
│ ┌──────┴──────┐ ┌──────┴──────┐ │
│ │ Analytics_ │ │ GMAT_ │ │
│ │ vis (.py) │ │ interface │ │
│ │ (Plotly) │ │ (.py) │ │
│ └─────────────┘ └─────────────┘ │
└─────────────────────────────────────┘
│
┌─────────────────▼──────────────────┐
│ OPTIONAL C++ BACKEND │
│ ┌─────────────────────────────┐ │
│ │ casas_cpp (pybind11) │ │
│ │ • RK4 / RK78 integrators │ │
│ │ • J2, drag, SRP models │ │
│ │ • STM (6×6) / STT (6×6×6) │ │
│ │ • EKF measurement update │ │
│ └─────────────────────────────┘ │
│ ▲ │
│ │ (fallback to Python) │
│ ┌──────┴──────┐ │
│ │ casas_ │ │
│ │ propagator │ │
│ │ (.py) │ │
│ └─────────────┘ │
└─────────────────────────────────────┘
🛠️ Project Structure & Missing Assets
If you are running the full enterprise dashboard system, verify that your local workspace includes these core components:
Code snippet
├── api_server.py # FastAPI server routing, payload checks & initialization
├── CDM_database.py # SQLite persistence engine with WAL support
├── Analytics_vis.py # Plotly data transformation & rendering routines
├── casas_propagator.py # Fallback layer managing native bindings
├── casas_da.hpp # C++ Core math, DA types, and integrator structures
├── bindings.cpp # Pybind11 registration declarations
├── build.sh # Automation compiler script for POSIX target environments
├── requirements.txt # Third-party dependency definitions
│
▼ REQUIRED FUNCTIONAL FILES (Verify these are present before starting):
├── CDM_parser.py # Interprets raw CSV streams into object definitions
├── Collision_Probability.py# Holds numerical risk assessment algorithms
├── Risk_Trend.py # Manages alert verification and event analysis
└── GMAT_interface.py # Exports orbital state arrays into NASA GMAT scripts
**
⚙️ Compilation & Environment Setup
1. Prerequisites
Ensure you have a C++17 compatible compiler installed (g++ or clang), alongside development headers for Python.**
**2. Install Dependencies**
pip install -r requirements.txt
**3. Compile the C++ Engine Backend
Execute the automated build script to construct and link the native binary extension:**
Bash
bash build.sh
**Note: This generates a compiled shared module file inside your root folder, allowing transparent performance improvements directly inside your Python runtime**.
**🖥️ Running the Application
To launch the backend API server locally, execute the server wrapper:**
Bash
python api_server.py
**By default, the system boots a worker instance accessible at http://127.0.0.1:8000. You can explore and test the interactive API endpoints directly through the automated Swagger documentation portal at /docs.**
Route,Method,Description
/,GET,HTML system dashboard landing view
/health,GET,System health check alongside record statistics
/api/events,GET,Fetches an aggregated list of close-approach events
/api/events/{id},GET,Retrieves historic updates for a specific conjunction
/api/events/{id}/trend,GET,Time-series risk evolution track
/api/events/{id}/pc,GET,Multi-method risk calculations & maneuver advice
/api/events/{id}/gmat,GET,Downloads an automated NASA GMAT script
/api/ingest,POST,Ingests a new CDM dataset into the database
/api/alerts,GET,Queries active system warning items
/api/alerts/{id}/ack,POST,Sets a warning message status to acknowledged