Full autonomous control of an assistive robot

Post links to your CoppeliaSim videos, scenes and models
Post Reply
Eric
Posts: 186
Joined: 11 Feb 2013, 16:39

Full autonomous control of an assistive robot

Post by Eric »

This video presents the desktop training platform that allows an operator to control a mobile robot in a smart environment. We show how an operator can navigate into a graphical user interface (GUI) (left screen) to define his desired location or action to do. Once defined the target action defined, the simulated robot (right screen) is autonomously actuated to perform the task.

Navigation in the menus: The operator is interacting with the GUI using a Brain Computer Interface (EEG or EMG) with a limited number of channels (i.e. few different means of interaction : next, previous, validate, cancel). Moreover,the usage of those channels is not very comfortable and can be mentally tiring due to classification errors, responding time, etc. For those reasons, the operator selects at first the line of the icon he wants to click, then navigates inside the line to click the desired icon. This allows less required interactions with the GUI.
In this video, the operator first selects a place to go, and decide to have a detour toward a point of interest (POI) (water in the video), along the way.

Control of the robot:
The robot s control is fully autonomous. Once a target location is defined by the operator via the GUI, the control algorithm use a topological path planner to define the succession of places to cross to reach the destination. Then for each of this places, it schedules actions to perform, to move from one place to the next one. Those actions are, in this video, using a metrical path planner to move toward the exit point, then move following a line on the ground (if any) until a Radio Frequency Identification (RFID) tag is detected (to reset the odometry and let the controller know the robot reached the next place). If along the way, an action at a POI is ordered by the operator, the action scheduler will include in his to do list, all the required set of sub actions to perform by the robot (e.g. for water: go to the POI location, get the water, let the operator drink, put back the cup).

http://youtu.be/ZHK6WYQvREU

Post Reply