Given the crucial role of spatial knowledge for life on earth, it is of fundamental importance to understand the rules and the mechanisms that have evolved for animals to be competent navigators, and in particular the active, anticipatory learning that is involved in acquiring navigational competence.
At their scale, insects are as good at this, as birds and mammals and we can learn much by understanding the basis of this competence in animals that are resource-limited with respect to sensors and computation. Visual navigation poses challenges for both biological and technical systems and it remains a stimulating and provocative fact that insects are smaller and more autonomous, more competent, more robust and more flexible than anything we can presently build.
We study insect navigation in the field, where we track ants with Differential GPS and record their scanning behaviour and the flight paths of ground-nesting bees and wasps with high-speed cameras. We also quantify the visual information content of navigation environments by analysing panoramic images. Image differences that are generated by translation can be used to indicate how well a location in space is identified by the view seen there and image differences generated by rotation can be used to investigate how well and where the compass direction of a reference image can be recovered.