Volley Electric Vehicle Prototype

Unfortunately, I am unable to share the code for this project as it is proprietary to Volley Automation. However, I can share the process and the technologies used in the project.

The Proto-Gantry project is a small-scale prototype designed to locate and approach an EV-charging adapter using a gantry system combined with an OAK machine vision camera. Key components include a YOLOv7 model for object detection, spatial processing scripts, and TCP-based gantry control.

Conceptual Techniques

Computer Vision with YOLOv7:

Skills include image dataset creation, annotation, and custom object detection model training using YOLOv7 on platforms like Google Colab, along with applying preprocessing and dataset management for object detection tasks.

Camera-Driven Object Localization:

Techniques for using an OAK camera to capture spatial data, detecting and tracking objects, calculating spatial coordinates, and translating those into actionable movements within a robotic gantry system.

Embedded System Control with TCP Communication:

Embedded programming to control motorized systems through TCP commands, emphasizing real-time communication, motor homing, and precision movements, along with implementing input parsing and error handling.

System Integration and Testing:

Skills in setting up hardware connections, addressing common challenges with motor responsiveness and network latency, and calibrating camera-based spatial detection within a physical range.