To build the core robotic functions of an autonomous vehicle system: sensor fusion, localization and control. All projects are implemented using C++ as programming language.
GitHub repository links to Term 2 projects are as given below.
1. Bi-cycle Tracking using Extended Kalman Filters (EKF)
Simulated lidar and radar measurements are provided to detect a bicycle that travels around the vehicle. EKF implementation uses the method called first order Taylor expansion to obtain linear approximation of the non-linear systems.
Curriculum / Expert Instructors: Mercedes-Benz Research & Development North America
Here is the link for the completed project - SDC-P6-Extended-Kalman-Filter
2. Bi-cycle Tracking using Unscented Kalman Filters (UKF)
As in project 1, simulated lidar and radar measurements are provided to detect a bicycle that travels around the vehicle. The Unscented Kalman Filter (UKF) is a novel development in the field. UKF implementation locates objects with better accuracy in highly non-linear systems.
Curriculum / Expert Instructors: Mercedes-Benz Research & Development North America
Here is the link for the completed project - SDC-P7-Unscented-Kalman-Filter
3. Kidnapped Vehicle
Implement a particle filter to take real-world data and localize a lost vehicle using,
- (noisy) GPS estimate of its initial location,
- lots of (noisy) sensor and control data.
Curriculum / Expert Instructors: Mercedes-Benz Research & Development North America
Here is the link for the completed project - SDC-P8-Particle-Filter
4. Lane Keeping - PID Controller Project
Implement a PID Controller to race around the lake track and to keep a simulated vehicle in its lane.
Here is the link for the completed project - SDC-P9-PID-Controller
5. Model Predictive Control
Implement a sophisticated control algorithm for stabilizing the vehicle in a noisy environment.
Curriculum / Expert Instructors: Uber ATG
Here is the link for the completed project - SDC-P10-MPC-Controller