The project was designed to contribute to the overarching aim of conceptually advancing biomimetics as a scientific discipline. In particular, it started from the question how biological systems manage the intrinsically contradictory aims of optimality with respect to specific functions on the one hand and multi-functionality and robustness on the other hand.
A particular methodological idea was to realize abstraction of the biological concept generator on the level of its formal mathematical description rather than its actual outer appearance. This means, that the biological system serving as a role model for a load bearing structure does not have to be a load bearing structure itself; it is not objectively similar to the technical system to be developed. Instead, the mathematical representation of the biological system ought to be similar to the mathematical representation of the load bearing structure, thus exploiting a /formal/ analogy rather than a /substantial/ analogy.
The project started from two fundamental working hypotheses:
- Hypothesis 1: Biomimetically designed load-bearing structures can in principle be even better than their biological concept generators.
- Hypothesis 2: The complexity of the structure of biological systems is positively correlated with robustness towards perturbations and imperfections.
Hypothesis 1 is derived from the reasoning that technical load bearing structures do not have to maintain as many different functions as natural systems and thus the aspect of optimality can benefit from the fact that fewer compromises in terms of multi-functionality have to be made. Hypothesis 2 has been concretely derived from recent works in the field of systems biology. One of the major disciplinary works in this field within this project was to verify or falsify this hypothesis by applying statistical approaches.
As a sample application, biological concept generators from systems biology have been used as role models. The relevant disciplinary research investigated signalling pathways in certain proteins and has been conducted by Prof. Nicole Radde at the Institute for Systems Theory and Automatic Control. As a brief summary, it can be concluded that a signalling cascade with higher complexity acts as an efficient low pass filter and is thus robust towards spurious variations in input parameters.
The first attempt was to transfer results for robustness of signalling pathways to load bearing structures on the basis of abstract matrix representations, namely so-called redundancy matrices in the field of structural mechanics. Indeed, these matrices can help to identify, localize and quantify robustness within engineering structures. However, a useful analogy to biological systems could not be found.
A more feasible formal analogy has been identified in the representation of dynamic systems via a set of differential equations that formally resembles the one that describes the aforementioned signalling cascades. In particular, it could be verified that both signalling cascades (systems biology) and truss structures (structural mechanics) share the feature that complexity is positively related to robustness, measured in terms of sensitivity towards perturbations or imperfections.
Structural design with biological methods: optimality, multi-functionality and robustness (B05)
German Research Foundation (DFG), CRC/Transregio SFB-TRR 141 "Biological Design and Integrative Structures. Analysis, Simulation and Implementation in Architecture ", GEPRIS project number 260977218
Institute for Systems Theory and Automatic Control (IST), University of Stuttgart