Purdue Grand Prix Simulator (KRP Custom Track Mod)

In connection with a long-term pilot study for improving driver training and proficiency for the Purdue Grand Prix, I created a highly realistic, drivable simulation of the Purdue Grand Prix track within Kart Racing Pro (KRP). The simulator was verified on many different racing-sim wheels (Logitech, DOYO, and PXN) and pedal configurations so as to mimic actual on-track driving conditions as closely as possible. It is a realistic practice and analysis platform for Purdue student drivers, especially those looking for extra practice time or high-level feedback experience prior to race weekend.

  • January 2024 – Present

    Ongoing research-based simulation project recreating the Purdue Grand Prix track using photogrammetry, Google Earth elevation data, and track modeling tools for driver training and kart telemetry. Built in Kart Racing Pro using Blender, 3ds Max, and Bob’s Track Builder.

    • Giancarlo Guccione – Lead Developer & Track Designer

    • Anthony Tan – Poster & Marketing Design (Research Conference)

My Role

I led the end-to-end design, development, and testing of the simulator. This included modeling the circuit from scratch, researching accurate topographic data, integrating performance telemetry tools, and facilitating feedback loops with student drivers. I also acted as the primary liaison between the project and Purdue Grand Prix teams, gathering user insights and coordinating future training applications. My role required a fusion of technical 3D design skills, game engine knowledge, motorsport fundamentals, and user-centered design thinking.

I also spoke with representatives from Williams Racing, such as Alexander Albon (F1 Driver) and Luke Browning (Academy Driver), for insights into driving realism elements, track characteristics, and sim-based performance analysis. Aided by such input, decisions about elevation complexity, surface reactivity, and optimal race lines were made.

My work needed to combine technical 3D design proficiency, game engine experience, motorsport basics, and user-centric design thinking.

Wireframe Overview (Bob's Track Builder)

Barrier Placement (Bob's Track Builder)

Grandstands Implementation

Track Development

The track was custom-built from scratch using 3Ds Max and Blender. I reconstructed the layout, terrain, pit lane, and barriers to reflect the physical Grand Prix course as closely as possible. A major focus of the project was achieving accurate elevation changes—an element often overlooked in amateur sim tracks.

Instead of relying on suspect onboard video footage, I used Google Earth’s “Show Elevation Profile” option to grab vertical data. Manually tracing an actual centerline path over the course enabled me to convert the elevation points one by one into the editor. A 1.61-meter elevation change section, for example, was precisely reproduced in the 3D model. It was one of the more reliable and cost-effective methods to achieve topographic accuracy on this scale.

Once elevation data was simulated, I imported the geometry into Bob’s Track Builder to determine track surface characteristics as well as object placements. Physics parameters were fine-tuned using 3dSimED with a provision for Kart Racing Pro compatibility for best mesh efficiency with maintained visual and functional verisimilitude.

GP Track Final Render 51 (Track POV)

GP Track Final Render 52 (Spectator POV)

Lights and Utility Implementation

AI Driver Performance Model and Race Strategy Engine

During Race 69 of the Purdue Grand Prix, a custom AI-driven analysis system was developed to transform raw telemetry into actionable performance insights. Built on top of Kart Racing Pro data, the model processes driver inputs in real time and post-session, identifying patterns in braking, throttle application, steering behavior, and speed across the track.

The AI analyzes sector-by-sector performance to detect where time is lost and why. It evaluates inconsistencies in driver input, highlights inefficient racing lines, and identifies suboptimal corner entry and exit behavior. Based on this, the system generates targeted recommendations, such as adjusting braking points, improving throttle pickup timing, and prioritizing exit speed in key sections.

Beyond driver input analysis, the model integrates kart setup data to begin forming a relationship between configuration and performance. By comparing telemetry across different runs, the AI can suggest setup adjustments and driving strategies tailored to specific track conditions and driver tendencies.

This system represents the foundation of a scalable race intelligence platform. As more real-world data is collected from Purdue Grand Prix teams, the AI will continue to improve in accuracy, with the long-term goal of delivering optimized race strategies, adaptive driver coaching, and data-driven performance tuning for competitive kart racing.

Impact

This project marks one of the first university-based, community-validated sim racing efforts for collegiate motorsport, fusing entertainment-grade platforms with simulation-level accuracy to better prepare student racers for real-world competition.

This simulation experience earned 2nd Place at the2025 Purdue Undergraduate Research Conference, recognized for its innovation in merging motorsport training and game development.

GP Track Final Render 50 (Start Line)

Final Map Test Run