Jeep ROS project

This respository was developed as the capstone project of Mechantronics Engineering 2019 at Tecnologico de Monterrey, Guadalajara, Mexico.

The objective of the project was to make a 2001 Jeep Cherokee sport follow a simulated street lane autonomously. Besides, stopping when a nearby object is detected.

This project was implemented and tested on NVIDIA JetsonTX2 running Ubuntu 16.04.

MCU Software

The development of the MCU (NXP S32k148) sofware layer could be found on Farid Izar repository.

YOLOV3 and ZED camera

We implement the well-known YOLOV3 algorithm to recognize objects as well as its distances with a ZED camera from StereoLabs.

• YoloV3 repository
• Zed Ros Wrapper

Author(s): Hector Cruz (Main developer), Ander Plascencia (lane_detector node), Farid Izar (Low-level Control)