Autonomous Navigation and Parking of a Robotic Car
Simulating a robotic car in Morse simulator
About
The main objective of this project is to drive the robotic car autonomously and park in the identified parking spot in the simulation environment while following the constraints. The main contribution lies in detecting static, dynamic objects and road lanes and creating a dynamic map based on projections for motion planning. The project work includes implementing an online prediction module using a RNN, tuning a Time Elastic Band (TEB) local planner, and implementing a finite state machine-based decision-maker to park the car autonomously.
Architecture
Software architecture
Results
Costmap layers
2D lane mapping
Evalution scene from the Morse environment
Decision making using finite-state machine
Team
Sumit Patidar, Alexandros Nikolaou, Sharang Kaul, Janis Freund, Jhorman Perez, Stoyan Karastoyanov