Fakultät Informatik

Bernstein Conference 2016 - Workshop

More for Less - Energy Efficiency in Neural Locomotion Control



 Philipp Stratmann,  Florian Röhrbein Alin Albu-Schäffer


21st September 2016


9am - 12.30 pm


Humboldt-Universität zu Berlin


Bernstein Conference 2016 Berlin

Presentation Slides:

PDF Slides (encrypted; please contact  organizers for the password)



The evolution of the central nervous system of limbed vertebrates has strongly been driven by the need to move with small nutritional requirements. While the brain still consumes 20% of the resting metabolic energy, the evolved neural motor circuits are highly optimized, as illustrated by the comparison to the controller hardware in bio-mimicking robots that typically consumes more energy than the mechanical actuators. To reduce muscular metabolic demands, the motor circuits adjust their control dynamically to changing mechanical environments. If we want to fully understand neural motor circuits, we therefore need to ask how gaits can be controlled with high efficiency.

This workshop will consider how the neural circuits in the cerebellum, brainstem, and spinal cord encode two aspects of efficient motor control: minimizing the metabolic costs of the neural calculations on the one hand and reducing muscular energy consumption by exploiting resonance properties of the biomechanical system on the other hand. To provide a broad overview over these two aspects, we will approach the topics both from a theoretical and experimental point of view. We invite speakers with a background in experimental and computational neuroscience that work on efficient movement control as well as engineers investigating the technological reproduction of the neural and biomechanical systems. 

We plan to provide an interdisciplinary investigation on the link between energy efficiency and neural locomotion control. Its evolutionary importance suggests that this link can provide fundamental insights and hypotheses about neural motor circuits. Besides its neuroscientific relevance, engineers working on bio-mimicking technology can benefit from these insights as a quid pro quo for their knowledge.






David Lester, University of Manchester

What is Neuromorphic Computing, and Why Does It Matter?


Christoph Richter, Technical University of Munich

Efficient Control of Musculoskeletal Robots with Spiking Neurons


Coffee Break


Florentin Wörgötter, Georg-August University Gottingen

Robots under Neural Control


Philipp Stratmann, Technical University of Munich / German Aerospace Center (DLR)

Low-Dimensional Control of Efficient Elastic Movements


Henrik Jörntell, Lund University

Towards an Integrative View of the Roles of the Spinal and Cerebellar Neuronal Circuitry in Brain Movement Control





Carola Jumpertz


+49 89-289-17552



no news in this list.