Stanley Heinze is an associate Professor in the Lund Vision Group at Lund University in southern Sweden. He started his group in 2015 after a two year postdoc in Lund and a preceding three year postdoc at the University of Massachusetts, USA. Originally from Germany, he received a Diploma degree in Biology at the University of Marburg, Germany, where he also carried out his PhD studies. Since his first degree he has been working on the neurophysiology and neuroanatomy of the insect central complex, especially in the context of navigation and orientation behaviors. His main focus of work at the moment is to compare the structure-function relationships within this central decision making center of the brain across a range of arthropod species and correlating species phylogeny, behavioral strategies and sensory environments to the neural circuitry of the central complex on the level of single neurons at synaptic resolution.
Using comparative connectomics of the insect central complex to guide the development of novel neuromorphic computers.
The abilities of insects to navigate are prime examples of how evolution has found elegant and efficient biological solutions to challenging problems. Insects have a comparably simple brain with a volume similar to that of a grain of rice and less than one million neurons. Yet, they master difficult tasks like returning to a well hidden nest in the dim, complex 3D maze of a jungle, or locating single mountain tops from hundreds or thousands of kilometers away, without having ever been there before. Some insects can establish optimal foraging routes between multiple food patches or can find shortcuts between known food locations. Independent of the strategy used to achieve that, navigation can be broken down in elementary navigational decisions. At each moment in time, an animal has to decide whether to turn right, left, or keep moving straight. To do that, its brain needs to compare the animal’s current heading with its goal direction and initiate steering commands in case the angles are not aligned. In insects, a brain region called the central complex achieves these tasks, independent of a species’ navigational strategy or sensory environment. We are thus asking how a single neural circuit can control the great variety of navigation behaviors found across the 1 million insect species on the planet. We have selected a representative set of species across the insect phylogeny and perform synaptic level connectomics of the circuits in the central complex. We have two aims, first identify the shared core circuits common to all insects, and second, identify how this circuit has evolved to solve specific navigational tasks of individual species. Both are providing answers required for adopting insect navigation circuits in robotic applications. To translate our biological data into novel hardware, we use computational models to extract the algorithms implemented in the insect central complex. Given tight structure function relations in this brain region, the geometry of these models is highly structured and has thus offered the possibility for implementation in a novel neuromorphic computing platform. Using nanowire based artificial neurons which communicate with light instead of electrons, we have conceived a design for a neural network that directly exploits circuit geometry for computations, eliminating the need for wiring, producing hardware with a tiny footprint, both in terms of physical size and energy consumption.