This paper presents a representation of the kinetic capabilities of a mobil
e robot, called the Dynamic Map. This representation is used to analyze int
eractions between robots and to plan trajectories according to a given task
. The representation is also used to plan the trajectory of a mobile robot
that must escape from several adversaries, as well as the collaborative int
erception of a robot by multiple opponents using the Dynamic Map concept. F
inally, the vision-based learning of a Dynamic Map is described. (C) 1999 E
lsevier Science Ltd. All rights reserved.