Interesting scientific controversies have been going for the last few years on recent trends in AI, sometime referred to as the "swing from symbolic to numeric AI". Such a dichotomy might not be very clarifying, nor does it seem very relevant for the advancement of the field. If one views and practices AI research as a multidisciplinary investigation whose purpose is the computational modeling and mechanization of a diversity of cognitive tasks (that may require embodiment in sensory-motor capabilities), then one has to face the challenges of integrating a diversity of mathematically heterogeneous representations. A single class of models highly adequate, for example, for data association can be totally ineffective for other purposes, such as, here, extracting and reasoning on the underlying ontology. The ambition of Integrative Heterogeneous AI is precisely to develop approaches and architectures capable of handling heterogeneous representations for complex tasks.

The presentation will motivate the challenges of Integrative AI and illustrate them on a very active research issue in the AI&Robotics community, which is the integration of motion and task planning. The basic techniques and representations for the two separate problems of task planning and motion planning will be briefly introduced, as well as the practical need for their integration. The main integrative approaches currently being explored will be surveyed.