- Real-vehicle-based Mid-to-Mid algorithm development
- Designing the mid-to-mid planning model architecture for real-world
- Performance validation through simulation and construction of machine learning based planning pipeline on real vehicles