🎯 BOW Benchmark

Rigorous benchmarking for Bayesian Optimization Approach to Windowing Motion Planning

📋 Overview

BOW Benchmark provides comprehensive performance comparisons between BOW (Bayesian Optimization Approach to Windowing Motion Planning) and state-of-the-art motion planning algorithms.

🚀 Supported Planners

Language	Planner	Description
C++	BOW	Bayesian Optimization Approach to Windowing Motion Planning
	RRT	Rapidly-exploring Random Trees (OMPL)
	HRVO	Hybrid Reciprocal Velocity Obstacles
	DWA	Dynamic Window Approach
Python	MPPI	Model Predictive Path Integral
	CBF	Control Barrier Functions

🛠️ Quick Start

First you have to install and setup the environment and then run.

Step 1: Install & Setup Environment

C++ Environment Setup

# Install dependencies
sudo apt update
sudo apt install build-essential cmake git pkg-config \
    libnlopt-dev libnlopt-cxx-dev libfcl-dev libompl-dev \
    libboost-all-dev libyaml-cpp-dev

# Clone and build
git clone <repo>
cd BOW
mkdir build && cd build
cmake ..
make -j$(nproc)

Python Environment Setup

# Setup conda environment
conda create -n bow python=3.10
conda activate bow
pip install -r requirements.txt

🔧 Anaconda Compatibility Fix

If you have Anaconda installed, add this to your ~/.bashrc:

export LD_LIBRARY_PATH=/usr/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH

Step 2: Run Benchmarks

C++ Benchmarks

# Run benchmarks
./build/benchmark/BowBenchmark

# Google Benchmark tests with custom options
./build/benchmark/BowBenchmark --benchmark_filter=BOW --benchmark_repetitions=10

Python Benchmarks

# Activate environment
conda activate bow

# Complete benchmark suite
./run_py.sh

# Custom experiments (edit config/main.yaml first)
python main.py

📦 Dependencies

C++ Requirements

Library	Purpose	Installation
NLopt	Nonlinear optimization	sudo apt install libnlopt-dev libnlopt-cxx-dev
FCL	Collision detection	sudo apt install libfcl-dev
OMPL	Motion planning	sudo apt install libompl-dev
Boost	C++ utilities	sudo apt install libboost-all-dev
YAML-CPP	Configuration parsing	sudo apt install libyaml-cpp-dev

Python Requirements

pip install -r requirements.txt

📁 Project Structure

BOW/
├── 📂 algos/                 # Python algorithm implementations
├── 📂 benchmark/             # C++ benchmark implementations
│   ├── 📂 include/          # Algorithm headers (dwa, hrvo, ompl)
│   └── 📂 src/              # Benchmark source code
├── 📂 bow/                   # BOW planner implementation
│   └── 📂 bow++/            # C++ BOW implementation
├── 📂 build/                 # C++ build directory
├── 📂 config/                # Configuration files
│   ├── 📂 planner/          # Planner configurations
│   └── 📂 test/             # Test configurations
├── 📂 include/               # C++ headers
│   ├── 📂 bow/              # BOW headers
│   └── 📂 limbo/            # Bayesian optimization library
├── 📂 results/               # Benchmark results
├── 📂 result_analysis_cpp/   # C++ result analysis
├── 📂 result_analysis_py/    # Python result analysis
├── 📂 scripts/               # Utility scripts
├── 📂 src/                   # Additional source implementations
└── 📂 test/                  # Test files

🔍 Verification

Check your installation:

# C++ dependencies
pkg-config --exists nlopt && echo "NLopt ✅" || echo "NLopt ❌"
find /usr -name "fcl" -type d 2>/dev/null | head -1 && echo "FCL ✅" || echo "FCL ❌"
find /usr -name "ompl" -type d 2>/dev/null | head -1 && echo "OMPL ✅" || echo "OMPL ❌"

# Python environment
conda activate bow
python -c "import numpy, matplotlib; print('Python environment ✅')"

🐛 Troubleshooting

Common Issues

GLIBCXX Version Mismatch (Anaconda)

export LD_LIBRARY_PATH=/usr/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH

CMake Configuration Issues

rm -rf build/ && mkdir build && cd build && cmake ..

Missing Headers

sudo apt install libnlopt-dev libfcl-dev libompl-dev libboost-all-dev libyaml-cpp-dev

Other Distributions

Fedora/RHEL/CentOS:

sudo dnf install nlopt-devel fcl-devel ompl-devel boost-devel yaml-cpp-devel

macOS (Homebrew):

brew install nlopt fcl ompl boost yaml-cpp

📊 Results & Publications

Visit our project website for detailed experimental results and publications:

🌐 bow-web.github.io

🤝 Contributing

We welcome contributions! Please ensure:

• C++ code follows project standards
• Python compatibility with 3.10+
• Proper benchmark integration
• Updated documentation

📄 License

[Add your license information here]

📚 Citation

If you use this benchmark in your research, please cite our work:

[Add your citation information here]

---

🌐 Project Website • 📧 Support • 🐛 Issues