Path Planning with the humanoid robot i Cub

Path Planning with the humanoid robot i. Cub Semester Project 2008 Pantelis Zotos Supervisor: Sarah Degallier 1 Biologically Inspired Robotics Group (BIRG)

Overview of the presentation � What is Path Planning? � Ant Colony Optimization � Performance Tests – Parameters selection � i. Cub robot – Steering � Integration of the ACO algorithm in the i. Cub simulation � Results � Future Improvements 2 Path Planning with the humanoid robot i. Cub

What is Path Planning? of the existence of a collision-free path in an environment with obstacles � Computation of such a path � Efficiency : to find the shortest path to the destination in short time. � Reliability : the robot must not collide with obstacles. � Examine 3 Path Planning with the humanoid robot i. Cub

Overview of the presentation � What is Path Planning? � Ant Colony Optimization � Performance Tests – Parameters selection � i. Cub robot – Steering � Integration of the ACO algorithm in the i. Cub simulation � Results � Future Improvements 4 Path Planning with the humanoid robot i. Cub

Ant Colony Optimization (1) �Biological inspiration: many ant species deposit on the ground a substance, called pheromone, on their way to the food. Other ants follow the path with the highest concentration in pheromone… �They will find a solution in any case, if it exists. Theoretical proof of the convergence to the optimal solution (Gutjahr 2000). 5 Path Planning with the humanoid robot i. Cub

Ant Colony Optimization (2) � Heuristic function: ACO algorithm is able to take advantage of the specific characteristics of a problem by using a well defined heuristic function. � In our implementation: the world considered as a grid of squares. � A set of artificial ants search for a short path from the start to the end. 6 Path Planning with the humanoid robot i. Cub

Ant Colony Optimization (3) � Each agent has a current position in the grid, can move by one square, 7 possible new positions. � At each step, random decision according to a probability distribution. • After all ants have found a solution, the shortest path is selected and is updated with pheromone. • Continue until no more improvement is happening or until a fixed number of iterations (generations). 7 Path Planning with the humanoid robot i. Cub

Overview of the presentation � What is Path Planning? � Ant Colony Optimization � Performance Tests – Parameters selection � i. Cub robot – Steering � Integration of the ACO algorithm in the i. Cub simulation � Results � Future Improvements 8 Path Planning with the humanoid robot i. Cub

Performance Tests – Parameters selection (1) � Heuristic function used: • Parameters to select: • Number of ants • Number of generations • α , the pheromone factor • β , the heuristic factor • 3 different environments (18 x 18) 9 Path Planning with the humanoid robot i. Cub

Performance Tests – Parameters selection (2) solution length, various number of ants 40 solution length 35 • 30 25 20 • 15 Average solution length of 10 executions 10 5 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 1 st environment. Increasing number of ants. 10 executions for each set of parameters. Average solution length starndard deviation ants standard deviation, 1 st environment 4 3. 5 3 2. 5 2 1. 5 1 0. 5 0 0 50 10 100 ants 150 200 250 • The variance of the found solutions decreases as the number of ants increases. • Stabilization after 100 ants. Path Planning with the humanoid robot i. Cub

Performance Tests – Parameters selection (3) solution length, various number of generations 70 solution length 60 50 40 30 Average solution length of 10 executions 20 Length of optimal solution 10 0 0 20 40 60 80 100 120 generations 140 160 180 200 220 solution length, various number of generations, 100 ants 70 solution length 60 50 40 30 20 Average solution length of 10 executions 10 0 0 20 11 40 60 80 100 120 140 generations 160 180 200 Path Planning with the humanoid robot i. Cub 220 • 2 nd environment. 20 and 100 ants. Increasing number of generations. • For 100 ants, stabilization after about 100 generations.

Performance Tests – Parameters selection (4) solution length, various number of generations, 100 ants 70 solution length 60 50 40 30 20 Average solution length of 10 executions 10 0 0 20 40 60 80 100 120 140 generations 160 180 200 220 solution length, various number of generations, best of 10 executions solution length 50 40 30 20 Best of 10 executions Length of optimal solution 10 0 0 20 40 60 80 100 120 140 160 180 200 generations 12 Path Planning with the humanoid robot i. Cub 220 • 2 nd environment. 100 ants. Increasing number of generations. • Comparison of average solution and best of 10 executions.

Performance Tests – Parameters selection (5) solution length, alpha-beta 100 80 11 12 13 14 60 40 20 solution length 80 solution length, alpha-beta 21 22 23 24 60 40 20 0 0 10 20 30 40 50 60 generations 70 80 90 100 solution length 0 generations • 3 rd environment. Different combinations of α and β. All combinations converge at about same quality solutions but in different number of generations. The pair α=2, β=1 was chosen. 13 Path Planning with the humanoid robot i. Cub

Performance Tests – Parameters selection (6) large values of α (bigger than 1), the algorithm is expected to stagnate to the first good solution found. � If a very good heuristic is available we should use bigger values for β in order to take advantage of it. � Environments with random distribution of obstacles : difficult to find a very good heuristic for every kind of problem. � Parameters selected for the integration of the ACO algorithm to the i. Cub: � Number of ants = 100. α = 2, β=1. � Number of generations : time limit or fixed number. � For 14 Path Planning with the humanoid robot i. Cub

Overview of the presentation � What is Path Planning? � Ant Colony Optimization � Performance Tests – Parameters selection � i. Cub robot – Steering � Integration of the ACO algorithm in the i. Cub simulation � Results � Future Improvements 15 Path Planning with the humanoid robot i. Cub

i. Cub robot – Steering (1) � i. Cub is an infant-like robot with the cognitive abilities of a 2 years-old child. � Its crawling is controlled by a CPG developed by Righetti (2006). � Add steering ability to i. Cub. Take advantage of the ability to turn its torso around 3 axises. 16 Path Planning with the humanoid robot i. Cub

i. Cub robot – Steering (2) � Rotate around y-axis and x-axis. • The outer limbs have to make a bigger step than the inner ones during steering. • Very small changes at each step, such that the motion to be as smooth as possible. 17 Path Planning with the humanoid robot i. Cub

i. Cub robot – Steering (2) 18 Path Planning with the humanoid robot i. Cub

Overview of the presentation � What is Path Planning? � Ant Colony Optimization � Performance Tests – Parameters selection � i. Cub robot – Steering � Integration of the ACO algorithm in the i. Cub simulation � Results � Future Improvements 19 Path Planning with the humanoid robot i. Cub

Integration of the ACO algorithm in the i. Cub simulation (1) � Board of 8 x 8 squares surrounded by a wall and with 5 square obstacles inside • 4 modifications. 1. The robot is not able to steer in a very big angle. 2. a diagonal movement must not be valid if there are two obstacles adjacent to the robot's movement. 20 Path Planning with the humanoid robot i. Cub

Integration of the ACO algorithm in the i. Cub simulation (2) � 4 3. 4. � 21 modifications. The robot prefers to move straight instead of steering. The robot prefers to move diagonally instead of making a 90º turn. The obstacles can change their position dynamically. Path Planning with the humanoid robot i. Cub

Overview of the presentation � What is Path Planning? � Ant Colony Optimization � Performance Tests – Parameters selection � i. Cub robot – Steering � Integration of the ACO algorithm in the i. Cub simulation � Results � Future Improvements 22 Path Planning with the humanoid robot i. Cub

Results • Grids, in which the i. Cub successfully reached the goal position. • Example of dynamically changed environment. The robot recomputes its plan to the goal position, when it detects a change in the positions of the obstacles. 23 Path Planning with the humanoid robot i. Cub

Overview of the presentation � What is Path Planning? � Ant Colony Optimization � Performance Tests – Parameters selection � i. Cub robot – Steering � Integration of the ACO algorithm in the i. Cub simulation � Results � Future Improvements 24 Path Planning with the humanoid robot i. Cub

Future Improvements � Recomputation of the robot's path to the goal position during the simulation, for instance after every 5 steps. � Multiple executions of the algorithm and selection of the shortest path. � Ability to detect if i. Cub is following a wrong route. � Integration of the ACO algorithm in the real i. Cub robot and test of its performance in real environment conditions. 25 Path Planning with the humanoid robot i. Cub

Questions 26 Path Planning with the humanoid robot i. Cub