BFS | DFS | A* | D* | Djisktra's

After creating a randomly generated obstacle field obeying density constraints, we had to implement and analyze the performance of a number of motion-planning algorithms, to generate paths from a start node to a goal node.

Hybrid A* | Dynamic Modelling | B-Spline | Djisktra's

Essentially a parallel parking system, a Hybrid A* algorithm is used to create a path from the source (top left) to the destination (parking spot) while taking into account sufficient space around the obstacles according to the vehicle dimensions.

Once paths are generated, dynamic considerations and smoothening are used to generate the final path ensuring that the heading angles do not exceed the maximum steering angle.

Hybrid A* | PRM | Path Smoothening | Dynamic Modelling | Resource Management

Motion planning for a non-holonomic firetruck in a randomly generated forest where new regions catch fire at regular time intervals. Two different planners were used to perform the simulation, and their performances are compared.

The ultimate goal is to optimally plan paths to the most important fire to fight, as fires spread, and new fires get started.

RRT | MATLAB | 3D Modelling

In a variation of the classical piano problem, we have to remove the main shaft from an engine transmission case without any collisions with the rest of the assembly.

A 6-degree bi-directional RRT algorithm is used to generate valid orientations that are not in collision, and a path is found to extract the shaft from the housing.