A smart driving simulator using CARLA and a Logitech G920 steering wheel trains drivers through realistic scenarios, real-time feedback, and crash playback to help correct common driving mistakes and improve road safety.
This project introduces an advanced smart driving simulator developed using the CARLA simulation environment and integrated with the Logitech G920 steering wheel to provide a realistic and interactive training experience. The simulator is designed to address persistent and dangerous driving habits such as tailgating, failing to check blind spots, and ignoring crosswalks—issues that often persist despite the availability of modern vehicle assistance technologies. By simulating real-world scenarios in a controlled, repeatable environment, the system allows drivers to experience the consequences of their mistakes and receive immediate, intuitive feedback.
The simulator features a variety of carefully designed scenarios that mirror common traffic violations and hazards. Key examples include making an unsafe left turn while other vehicles have the right of way, failing to yield to pedestrians hidden by other cars at a crosswalk, and following vehicles too closely, which can lead to rear-end collisions. These scenarios are enhanced with dynamic environmental elements such as AI-controlled vehicles, pedestrians, traffic lights, and changing weather conditions, all managed via CARLA’s powerful traffic and physics engines.
To guide and educate drivers during the simulation, the system includes directional arrows, speed displays, turn indicators, mirrors, and multi-angle camera views (rear, left, right, and overhead). Upon collision, the simulator records the last few seconds before and after the impact and offers replay functionality from multiple viewpoints to help drivers review their actions and learn from their errors.
In addition to its educational value, the simulator features a user-friendly interface with an opening screen, instruction pages, and a feedback-based ending screen, making it accessible for a wide range of users. The project ultimately aims to improve driver behavior and road safety by offering an engaging, personalized, and data-driven learning experience, with future potential for expansion into professional driver training and autonomous vehicle testing.
