Mechatronics engineer | Software developer
Planning and control for vehicles that have to work.
PhD candidate at the University of Trento. I build motion planners, optimal-control solvers and real-time C++ for autonomous driving and motorsport — from sampling-based tree search to minimum-lap-time simulation. Most of what I publish here is numerics I'd rather own than depend on.
🌐 mattiapzz.github.io · 💼 LinkedIn · 📄 Scholar · 🆔 ORCID
- Motion planning — sampling-based planners (MPTree / MPTreeRRT, MCTS), vehicle motion primitives, clothoid-based geometry
- Optimal control — minimum-lap-time simulation, NLP solvers, symbolic model generation
- Real-time C++ — no-alloc, embedded-friendly libraries meant to run on a vehicle, not in a notebook
| Project | What it is |
|---|---|
| FBGA · paper | Forward-backward time-optimal velocity planning under generic acceleration constraints (C++). FBGA_3D extends it to 3D spatial paths with a generic GGGV envelope. |
| VehiclePrimitives · paper | Motion primitives and the MPTreeRRT planner — feasible trajectories for cars, trucks and mobile robots. |
| geographic_coordinates | WGS84 ↔ ECEF ↔ ENU conversions. Turning raw GNSS fixes into a usable local frame. |
| scope | Zero-dependency browser scope for telemetry CSV/TSV — series, XY, XYZ. No build step. |
| CarDashboard-Sim | Live vehicle dashboard emulator in plain HTML. Demo |
Institutional work — VeDi 2025 (CRF/Stellantis), SUNRISE (Horizon Europe), Minimum Lap Time Simulator (Aprilia Racing) — is described on my site.
Daily: C/C++ · MATLAB · Maple · CMake · Git Regular: Python · TeX · Quarto · Docker · ROS · CarMaker · Mathematica Occasional: Ruby · Lua · R Learning: Rust, Zig, JavaScript
Open to conversations about planning, control, vehicle dynamics, and racing — feel free to reach out.