Automated wireless power transfer can be seen as an enabling technology for autonomousvehicles. At SINTEF Energy Research a 1/14 scale electric truck model  had beenused for demonstrating such technology. The truck was fitted with an induction system,but was otherwise stock. In this master thesis a system for autonomous path following hasbeen developed and successfully implemented to demonstrate self-driving capabilities ona path that includes an inductive charger. The purpose of the vehicle is to visualize theconcept of wireless charging for a self-driving truck, but on a smaller scale.
Three different methods has been tested and implemented.
1. A LiDAR-based approach that utilizes Hector SLAM for mapping and localizationtogether with a Pure Pursuit and Stanley Steering path tracking steering controller.
2. A machine-learning and camera-based approach that can navigate a track using deeplearning and behavior cloning/supervised learning. This was inspired by the convolutionalneural network used in Nvidia’s DAVE-2 self-driving car.
3. A classic computer vision approach for detecting lane curvature and using PID forcontrol.
ROS nodes that integrates the output from the SLAM- and deep learning-methods andcombines them with a path recorder and a path tracking algorithm has been made, togetherwith a graphical interface for monitoring vehicle states. Methods for estimating odometrywithout wheel encoders, dynamic speed control and obstacle detection with start/stopfunctionality has been implemented, along with a computer vision method for detectingArUco markers. A complete, self-driving ROS-based system has been created. An NvidiaJetson TX2 has been used as the main embedded computing unit. The system implementsHector SLAM as the primary SLAM method, since it does not require odometry. ThePure Pursuit path tracker was chosen as the preferred steering controller since it resultedin smooth and stable tracking of the path. The neural network of the machine-learningapproach was able to steer the truck reliably and is implemented in the final version asa separate mode. This works independently of the LiDAR for when the LiDAR is outof range. The neural network has also been tested on the Udacity self-driving simulator. The computer vision approach was found to be too demanding for the embeddedhardware, and ended up not being used. The inductive charger system has successfullybeen connected and integrated to the Jetson TX2 through a CAN-bus interface in orderto receive the battery state. The amount of charge received is dependent on the positioningof the truck, and is not optimized for a system with manual path creation. A betterautonomous positioning system when driving across the charger is recommended, alongwith a better and more precise LiDAR and vehicle platform.