Hydrakon is the full stack autonomous system developed by Atlas Racing, The Formula Student team from Heriot Watt University Dubai.
Hydrakon is a modular, ROS2 based architecture that enables real-time autonomous driving in both simulations (Carla) and on the physical track (ADS-DV). Our system is designed for the Formula Student AI/Driverless competition and research in autonomous navigration, planning, and control
- Perception: Cone detection using camera (depth and mono) only
- Path Planning: Adaptive Pure Pursuit with turn classification and dynamic widening
- Control: Working on implementing PID and MPC controllers
- Vehicle Interface: CAN-based communication for interfacing directly with the ADS-DV
- Hydrakon V1 was developed as part of our participation in the FS UK 2025, Below are the code bases for Hydrakon V1. These repositories are not maintained anymore.
- HydrakonV1 houses the code we used to control the ADS-DV directly
- HydrakonSimV1 houses the code we used to build and develop our simulation in Carla
- At Formula Student UK 2025, Hydrakon powered Atlas Racing to:
- Real World AI Award Winners
- Babcock Overall Statics Award Winners
- First GCC-based team to claim the stage at Formula Student
Engineered with precision. Powered by ambition. Built for the future
