Abstract
We consider the problem of terrain model acquisition by a roving point placed in an unknown terrain populated by stationary polyhedral obstacles in two/three dimensions. The motivation for this problem is that after the terrain model is completely acquired, navigation from a source point to a destination point can be achieved along the collision-free paths. And this can be done without the usage of sensors by applying the existing techniques for the well-known find-path problem. In this communication, the Point Robot Autonomous Machine (PRAM) is used as a simplified abstract model for real-life roving robots. We presen! an algorithm that enables PRAM to autonomously acquire the model of an unexplored obstacle terrain composed of an unknown number of polyhedral obstacles in two/three dimensions. In our method, PRAM undertakes a systematic exploration of the obstacle terrain with its sensor that detects all the edges and vertices visible from the present location, and builds the complete obstacle terrain model.
| Original language | English |
|---|---|
| Pages (from-to) | 450-455 |
| Number of pages | 6 |
| Journal | IEEE Journal on Robotics and Automation |
| Volume | 4 |
| Issue number | 4 |
| DOIs | |
| State | Published - Aug 1988 |
| Externally published | Yes |
Funding
Another interesting problem ia the navigation of a robot in an unexplored or a partially explored terrain. In this case, the entire terrain model may not be known, and the robot relies on its sensors for navigation. Lumelsky and Stepanov 141 present sensor-based navigation algorithms for navigating a point automaton to a destination point using “touch” type of sensor. In this inethod localized sensor information is used to guide the point automaton, and this information is not put to any further global use. In many applications, incidental learning is shown to be an important enhancement in the navigation planning. Here, a composite model of the terrain is built by integrating the sensor information obtained as the robot executes sensor-based and goal-directed navigation. Iyengar et al. [2], Oommen et al. [6], Turchan and Wong 191 discuss different versions of learned navigation in unexplored terrains. Here we consider the problem of acquiring the terrain model by systematic exploration of the terrain using a sensor. Our main motivation stems from the fact that the availability of the terrain model enables us to plan the entire Manuscript received February 26, 1987; revised December 7. 1987. A preleminary version of this paper was presented at the 3rd IEEE Conference on AI Applications, Orlando, FL. Feb. 1987. The work of B. J. Oomman was partially supported by the National Sciences and Engineering Council of Canada. N. S. V. Rao is with the Department of Computer Science, Old Dominion University, Norfolk, VA 23529-0162. S. S. Iyengar is with the Department of Com:)uter Science, Louisiana State University, Baton Rouge, L.4 70803. B. J. Oommen is with the School of Computer Science. Carleton University, Ottawa KIS SB6, Canada. R. J. Kashyap is with the Department of Electrical Engineering. Purdue University, West Lafayette, IN 47907. IEEE Log Number 8820163.
| Funders | Funder number |
|---|---|
| National Sciences and Engineering Council of Canada |