Intelligent Machines? – Self-organized Nonlinear Dynamics of Machines across Scales

Workshop Report

The international workshop InteMa22 took place at the MPIPKS from June 20-24, 2022. The aim of the event was to bring together researchers from nonlinear dynamics and complex systems science with those from smart systems’ and robotic engineering and artificial intelligence to further our understanding of the dynamics of complex multi-dimensional machines – specifically those that themselves shall exhibit forms of smartness or intelligence. The workshop had 30 participants working in the areas of mechanical systems, smart materials, neural networks and learning systems as well as complex information processing and communication systems and concepts.

Investigating the ability to efficiently move and interact with complex environments was one of the major themes discussed at the workshop. Auke Jan Ijspeert presented a series of models demonstrating how robustness to neural disruptions can be achieved in robotic locomotion. Using machine learning for adaptive interlimb coordination in bio-inspired robots was the topic of Poramate Manoonpong’s talk, while Bulcsu Sandor offered insights into creating a network of neural oscillators for generating hexapod-type motion patterns. Maziar Ahmad Sharbafi and Philipp Rothemund explained solutions aiming in improving robot/smart device interactions with the environment. Contributions to the topic were made also by Je Min Hwangbo, Thies H. Büscher, Claudius Gros and Xiaofeng Xiong. Short-term, few-shot learning and the mechanisms underlying the learning dynamics has been identified as one of the overarching open topical questions.

A second most relevant topic was (self-)organization in multi-agent systems. At the workshop’s colloquium, Metin Sitti presented a self-organizing robotic collective system at the microscale capable of having interactions inducing various nonlinear coupled dynamics and dynamic self-assembly modes. Christian Bettstetter discussed in his talk a proof of concept for a multi-robot system of small ground robots and drones that forms emergent space-time patterns. Self-organized distributed patterns in the collective response dynamics of fluctuation-driven power grids was the topic of Xiaozhu Zhang’s talk. The many different aspects of the workshop’s topic were reflected in the talks’ diversity: smart materials (Thomas Wallmersperger, Zhenwu Wang), smart particles (Holger Stark), new approaches to computing (Lucas Wetzel, Wei Lu, Yulia Sandamirskaya), apps for sustainability awareness (Johannes Klinglmayr), etc.

The workshop connected many established researchers with junior scientists. Participants actively engaged with each other’s research and several repeatedly brainstormed, debated, or worked on open challenges after talks.

Despite the event taking place in a hybrid format, the workshop succeeded in giving an impulse for collaborative research. We would like to thank the MPIPKS for providing the opportunity for this highly relevant cross-disciplinary exchange, Maria Voigt for providing dedicated and outstanding organizational support and Ronny Börner for performing miracles on all IT tasks during the workshop. Thank You!