The main script to run is MAIN_Universal.m
% >>> The code calculates an A* path oblivious to the geometric constraints % of polyhedra using code adapted from "Interactive A* search demo % 04-26-2005, Copyright 2009-2010 The MathWorks, Inc." % After an A* path is generated, moves are considered that roll along and % try to adhere to this optimal A* path, with geometric constraints. % The code proceeds to calculate the footprint of the polyhedra, i.e. the % current downface and 2D net of possible next rolls as it % rolls along, considering the presence of obstacles, and tries to get it % within a specified distance to the goal point by connecting together % the footprints as it goes.
% >>> Note actual convergence on the goal point can % be impossible; it depends on the placement of the point and the user's % selected side length for the shape, etc. You define side length in the % main while loop condition: 6 < dist_to_goal. % There are problems with this naiive implementation of path planning; % it looks only 1 step ahead each time. It can get stuck in cycles, and the % path it generates can be far from optimal. If you have such problems, try % tuning the sideLength, tolerance, and pathSegments variables below. % Even small changes can impact convergence.
% From "Robert Baines, Joran Booth, and Rebecca Kramer-Bottiglio, Soft % rolling membrane-driven tensegrity robots. RAL. 2020" % https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9162439