Sharon StansfieldAssociate Professor |
The ultimate goal of this research is to explore the use of robots as assistive technologies for people with mobility impairments. In order for a robot to function in an environment such as the home, it must have the ability to sense and remember its world and to use this data to move around and carry out tasks. This is a very difficulty problem if the robot is to be fully autonomous. In an assistive mode, some of the more complex reasoning about how to do things may be left to the user. To do this, the user must have a reasonable, non-technical way to command and control the robot. This project addressed the beginning phase of such a system: building a model of the environment and visualizing the robot’s position and sensor data.
The robot used for the work is a Pioneer 3 mobile robot with a ring of eight sonar sensors in front and another ring of eight sonar in the back. The ARIA software library was used for reading the sonar sensors and the robot’s position and for controlling the robot. The 3DGameStudio engine was used for the real-time visualization of the robot data. 3DStudioMax modeling software was used to develop the model of the robot.
Software was developed to permit the robot to wander in its real environment while taking sonar readings. These readings are used to determine if there is an obstacle in front of each of the robot’s sonars and to get a measurement of the distance to that obstacle. As the robot moves in its world, this information is used to build an “occupancy grid” of the real world environment. This grid is essentially a 2D mapping of the sonar data that indicates whether the robot believes there is an obstacle at that position in its world. This data, along with the position of the robot was then sent to the visualization software, where it was displayed as a grid of cubes of different shades of blue. The darker the cube at a given position, the more confident the robot is that there is something there in the real world. At the same time, the position and movement of the robot is also displayed in the virtual world, so that the user sees a graphical representation of how the robot is moving in its environment. A customized C++ application was developed to handle the communication of data from the robot and to drive the updates of the graphical world used for visualization.
