How does a 0.1-mg brain solve a complex computational task? This is the question we are aiming towards answering by using the Saharan desert ant Cataglyphis as a model system for studying navigation within 3-D space, and by combining neurophysiological, behavioural, computational and finally robotics approaches. Central questions of our project are:

(i) whether ants do reduce a 3-D orientation task to a 2-D problem, or whether they perform path integration in a true 3-D manner,

(ii) how different kinds of sensory (proprioceptive) input interact to enable error-tolerant orientation in 3-D space, and

(iii) whether path integration finally enables a map-like cognitive behaviour.

In general, we try to understand how different modes of spatial orientation can interact collectively and set the stage for intelligent behaviour to arise. Conceptually speaking, the small-brain navigator Cataglyphis might provide an object lesson for studies in cognition and artificial intelligence, especially in attempts to understand how complex (high-level) behaviour is achieved by the interaction of rather simple (low-level) subroutines.