Trajectory Localisation Analysis Tool

A comprehensive tool for analysing localisation performance in ROS2 systems. This tool processes ROS2 bags containing pose data, performs trajectory analysis, and provides detailed metrics and visualisations.

Features

• Bag Processing

  • Automatic segmentation of long trajectories
  • Support for PoseStamped messages
  • Configurable segment duration
  • Robust pose count validation

• Trajectory Analysis

  • Absolute Trajectory Error (ATE) calculation
  • Relative Pose Error (RPE) metrics
  • Scale drift detection
  • Rotation error analysis

• Visualisation

  • 3D trajectory plots
  • Error heat maps
  • Time-series analysis
  • Comparative visualisations

• Dashboard (TODO)

  • Real-time metric monitoring
  • Customisable plots
  • Segment comparison
  • Export capabilities

Prerequisites

• Docker >= 20.10
• Docker Compose >= 2.23
• Git

Installation

1. Clone the repository:

   git clone https://github.com/usamahz/trajectory-analyzer.git #Skip this step if you already have the repository cloned
   cd trajectory-analyzer

2. Build the Docker containers:

   docker-compose build

Usage

1. Input Data:

   • Place your ROS2 bag file in data/input/ along with metadata file data/input/metadata.json

2. Run Analysis:

   ./run_analysis.sh

3. View Results: Analysis outputs can be found in the following directories:

   • data/output/analysis_summary.json - Overall analysis metrics
   • data/output/segment_*/plots/ - Visualisation plots for each segment

   Note: Interactive dashboard viewing (via `launch_dashboard.sh`) is placeholder and not implemented yet.

Configuration

The tool can be configured through several YAML files:

• Analysis Configuration (config/default.yaml):
  • Segment duration
  • Trajectory association parameters
  • Logging settings

Output Structure

Analysis results are stored in data/output/ with the following structure:

• analysis_summary.json: Overall metrics
• segment_X/: Individual segment analysis
  • poses/: Trajectory data
  • plots/: Visualisation plots
  • metrics/: Detailed metrics

CI/CD Workflow

The project proposes a comprehensive CI/CD pipeline using modern DevOps tools and practices:

Tools and Infrastructure

1. Version Control

   • GitHub for repository and branch management

2. CI/CD Orchestration

   • GitHub Actions for automating workflows, builds, and deployments

3. Containerization

   • Docker for creating portable, reproducible runtime environments

4. Infrastructure Automation

   • Terraform for automated provisioning of cloud resources (AWS EC2, S3)

5. Artifact Management

   • AWS S3 for storing build artifacts, intermediate results, and reports

6. Testing Frameworks

   • Pytest for unit tests, integration tests, and SLAM metrics validation

7. Data Processing Frameworks

   • Pandas for preprocessing trajectory data
   • ROS2 libraries for bag file parsing and trajectory evaluation

8. Monitoring and Reporting

   • Streamlit / Prometheus + Grafana for real-time pipeline monitoring
   • Slack for CI/CD event notifications (build status, failures, nightly runs)