Self Driving Car Engineer

Welcome to the Self-Driving Car Engineer Nanodegree program! Learn about the Nanodegree experience, as well as hear from Waymo, one of Udacity's partners for the program.

In this course, you will develop critical Machine Learning skills that are commonly leveraged in autonomous vehicle engineering. You will learn about the life cycle of a Machine Learning project, from framing the problem and choosing metrics to training and improving models. This course will focus on the camera sensor and you will learn how to process raw digital images before feeding them into different algorithms, such as neural networks. You will build convolutional neural networks using TensorFlow and learn how to classify and detect objects in images. With this course, you will be exposed to the whole Machine Learning workflow and get a good understanding of the work of a Machine Learning Engineer and how it translates to the autonomous vehicle context.

Besides cameras, self-driving cars rely on other sensors with complementary measurement principles to improve robustness and reliability, using sensor fusion. You will learn about the lidar sensor, different lidar types, and relevant criteria for sensor selection. Also, you will learn how to detect objects in a 3D lidar point cloud using a deep-learning approach, and then evaluate detection performance using a set of metrics. In the second half of the course, you will learn how to fuse camera and lidar detections and track objects over time with an Extended Kalman Filter. You will get hands-on experience with multi-target tracking, where you will initialize, update and delete tracks, assign measurements to tracks with data association techniques, and manage several tracks simultaneously.

In this course, you will learn all about robotic localization, from one-dimensional motion models up to using three-dimensional point cloud maps obtained from lidar sensors. You’ll begin by learning about the bicycle motion model, an approach to use simple motion to estimate location at the next time step, before gathering sensor data. Then, you’ll move onto using Markov localization in order to do 1D object tracking. From there, you will learn how to implement two scan matching algorithms, Iterative Closest Point (ICP) and Normal Distributions Transform (NDP), which work with 2D and 3D data. Finally, you will utilize these scan matching algorithms in the Point Cloud Library (PCL) to localize a simulated car with lidar sensing, using a 3D point cloud map obtained from the CARLA simulator.