| Abstract: | In this paper, the problem of path planning for sample collection using single or multiple mobile robots such as Mars rovers is formulated as a mathematical optimization problem involving a performance metric based on the scientific values of the collected rock and soil samples. The posed optimization problem is NP‐hard and more complex than the well‐known Traveling Salesman Problem. Algorithms are proposed for obtaining near‐optimal solutions for both single and multiple robots. Their application is illustrated using real Mars surface data. The dependence of the optimal performance on the number of mobile robots is studied numerically. © 2004 Wiley Periodicals, Inc. |
