Neurobiology

Research Interests

• Neuronal mechanisms underlying visually guided behaviour in insects
• Spatial learning and navigation
• Information processing by nerve cells and neuronal networks
• Population coding of natural sensory information
• Virtual reality techniques to analyse neuronal coding of natural visual information
• Active vision strategies in the context of spatial learning and navigation
• Neuromodelling and bio-inspired robotics

• Modelling neural mechanisms using numerical model simulations to test the plausibility of experimentally established hypotheses. Since the neural mechanisms of behavioral control work in closed feedback loops, a seemingly simple mechanism can show unexpected emergent properties in a closed loop situation.

• Transfer of bio-inspired algorithms for visual course-control and navigation to mobile robots. Insects solve complex tasks of behavioral control with their tiny brains. In many respects current robots are far less perfected. Therefore it is promising to emulate biological approaches to these tasks to improve mobile robots.

• Development of virtual reality systems for walking and fast flying insects. The highly specialised sensory and motor systems of these animals necessitate special hard -and software for appropriate animal- machine interfaces. One task is the generation and presentation of realistic panoramic images at high frame rates, another task is the measurement of motor actions of the animal and the realtime feedback to the image generation.

• mechanisms of multi-guidance in human navigation
• sensory cue integration
• distance-dependent cue weighting
• landmark usage and path integration

Hymenopterans, such as bees and ants, travel repeatedly during the course of the day between their nests and different food sites. They navigate in a wide range of environments, from deserts (e.g the ants of the genus Cataglyphis) to tropical forests (e.g. the bee Euglossa bidentata), and on very different spatial scales, from few meters (e.g. pinpointing the nest entrance) to several kilometers, i.e. the foraging range of the animals. During their trips, they have to steer toward their goal (e.g. their nest), but on their way they should not collide with obstacles.
The spatial information relevant for navigation on the wide range of relevant spatial scales and under variable environmental conditions may differ tremendously. How do hymenopterans acquire the relevant information? During large-scale navigation, hymenopterans partly rely on determining the direction of the goal, e.g. by using a compass, and the length of the traveled distance, e.g. provided by visual odometry in flying hymenopterans or by step counting in the case of ants. However, egocentric cues are prone to cumulative error. Therefore, additional mechanisms based on allocentric cues are necessary and have been shown to be used for small-scale (local) navigation.
The environment is rich of allocentric information. It contains odors, sounds, textures of objects, the distances to objects, etc. A single view of the scenery may contain several objects that are defined by cues such as contrast, their edges, surface texture or color. Moreover, the relative distance to objects can be extracted from the image motion generated by self-motion of the animal. However, despite the large number of studies on navigation in hymenopterans, it is still controversial which environmental features are used for naviagion and how they are combined.
Extracting relevant information is not the only problem for navigation. Indeed this information has to be stored and recalled at the relevant time. When a bee or wasp departs from an important place, such as its nest or a food source, it may perform, depending on its familiarity with the site, peculiar flight sequences ('learning flights') that have been concluded to help, on the one hand, in gathering visual information about the environment in the vicinity of this place, and, on the other hand, to memorize the information and action to perform to reach the goal again. The memory then needs to be correctly recalled during the return flight. How hymenopterans memorize and recall the information is still a largely open question.
Navigation is moreover tightly linked to motor control. What are the necessary and sufficient motor control strategies to navigate with a given type of information? Hymenopterans need, for instance, to decelerate before entering the nest, or when the environmental features used for navigation require time to be extracted. They also need to pay attention to changes in the environment, such as a moving car, not to collide with the objects. Therefore the motor control needs to integrate the information from the navigation algorithm and the collision avoidance algorithm.