Racing
Engineering.
From FSAE to IndyCar speeds — vehicle simulation, dynamics modelling.
IMS · Sep. 2024
Champ.
2022
Las Vegas 2024
CAVALIER AUTONOMOUS RACING
296 KM/H · WORLD RECORD
Cavalier
Autonomous
Racing
World record 296 km/h · IMS Sep. 2024
Predicting
at the limit
At IndyCar speeds, physical models accumulate systematic errors at the lateral limits that affect trajectory planning accuracy.
Dual-Track
MATLAB Sim
Dual-track MATLAB model for the Dallara AV-21. STM/DTM comparison, Pacejka tyre forces, load transfer. Telemetry pipeline for IAC data validation.
IAC
Events
| Event | Result |
|---|---|
| IAC@CES Las Vegas | 2nd Place |
| IAC IMS Time Trials | 1st · 296 km/h ↑ |
Prediction
Improvement
DTM captures load transfer and tyre nonlinearity that STM misses at the lateral limit. Evaluated against IAC telemetry.
| Metric | STM | DTM |
|---|---|---|
| RMSE aᵧ ↓ | — | — ↓ |
| RMSE v ↓ | — | — ↓ |
| Max error ↓ | — | — ↓ |
| Δ vs. STM | Ref. | −XX% |
FSAE
Racing
Two FSAE teams — design engineer at GETracing Dortmund building cockpit ergonomics, then Head of Sponsoring at GreenTeam Stuttgart during the world championship season.
Champion.
GreenTeam
Uni Stuttgart
- —Sponsorship strategy — high six-figure funding secured
- —Led 30+ member sponsor relations team
- —4 wins in 5 major European FSAE events
GETracing
FSAE
- —CAD redesign of dashboard, steering wheel and shift paddles in Creo 5
- —−70% weight via topology optimisation
- —Driver ergonomics validated via real-driver fit testing
Vehicle
Dynamics
Tool
The simulation framework built alongside the race engineering work. STM, DTM, Pacejka MF, GGV envelope, QSS lap simulation. No institutional context — built out of pure intrinsic motivation.
Modular
by Design
Chain of independently validatable modules. STM → DTM, Pacejka MF, brush model, load transfer, GGV, QSS lap sim. GUI has no physics backend of its own.
VDT Browser GUI
GUI only — no physics backend yet. Parameter sweeps and real-time plot visualisation.
Design
Projects
Motorsport bodywork, cockpit ergonomics, and surface design. F1 sidepod external aero study in CATIA V6. FSAE cockpit redesign in Creo 5 — −70% weight. Roboracer surface model at PoliMi.
External
Aero Surface
Full F1 sidepod surface geometry — inlet, undercut, and fin. Final project of the MEA CAD CAP course. Advanced Class-A surfacing workflow in CATIA V6.
Creo 5
−70% Weight
Full CAD redesign of dashboard, steering wheel, and shift paddles for GETracing FSAE. −70% weight reduction via topology optimisation. Validated through real-driver fit testing.
PoliMi
Surface Model
Surface model produced at Politecnico di Milano as part of the Surface Design for Engineering Applications course. Advanced Class-A surface modelling for the PoliMi Roboracer autonomous racing platform.
F1TENTH
Sim Race
Full autonomous stack for 1/10-scale racing. Builds on IAC architecture. Extends TUM F1tenth stack with own tyre model integration, DTM-based MPC controller, and simulation-first development pipeline.
TUM Stack ↗ GitHubFrom IAC
to F1TENTH
Porting thesis vehicle models into a ROS2 autonomous stack for 1/10-scale sim racing.
Scale gap:
AV-21 to 1/10
IAC experience at 296 km/h doesn't transfer directly to 1/10-scale racing. Standard stacks use kinematic models — too simple to race at the friction limit.
DTM-MPC
in ROS2
Port thesis STM/DTM into F1TENTH ROS2 stack as MPC plant model. Full pipeline: LiDAR → state estimation → DTM-MPC → control output. Simulation-first, hardware second.
ICRA 2026
in development
Race
Performance
DTM-MPC vs. kinematic baseline on the ICRA qualification map. Physics-informed plant model shows faster lap time convergence and higher top speed at the friction limit.
Racing
Career
Apr 2024
May 2022
Sep 2019
Jul 2023
Universities
↑ here
Present
Sep 2022
Predictive
State
Estimation
Comparing physics-based, data-driven, and hybrid vehicle models for real-time-capable predictive state estimation. The DTM foundations from IAC simulation — extended with machine learning.