Fall Validation Demo

Transport of Modular Payload using Autonomous Vehicle – Simulation

Autonomous Alignment and Docking with a Modular Payload with self-correction: Simulation

Verifying Docking Error Using 3D Point Cloud

Autonomous Docking (PR11)

Spring Validation Demonstration 2020

Project Pictures from Fall 2020

The pictures from Fall 2020 describe our attempts to set up the hardware we received from PIX Moving (fitting and installation of wheels on rims, trying to fix a discharged Li-Ion battery pack), as well as various simulation components that we developed, including the vector map, GUI, the docking verification visualization in RViz, various types of paths being planned and executed, and the realistic pod and chassis being simulated in Gazebo. In general, these images complement the ones from the Spring 2020 semester, since we built upon most of the functionality developed during that period.

Project Pictures from Spring 2020

Some of these pictures describe our hardware system development (the RC car and the camera, the pointcloud map of NSH B-level generated with the Velodyne VLP-16 LIDAR), whereas some of these detail our software and simulation system development, like the AprilTag detection, and the Carla and Gazebo simulations for localization and point-to-point navigation. Finally, the last few images show our custom docking simulation for approach navigation in Gazebo, where we tried out reinforcement learning and relatively complex maneuvers like Dubin’s path generation for planning before settling on pure pursuit. The final image shows the ideal goal we have of creating a modular pod and autonomous chassis system that can move offices, gyms and restaurants to anyone, anytime, anywhere.