Methods and Algorithms for Scalable Systematic Biologically-inspired Design

Biologically-Inspired Design is a specific type of between domain Design-by-Analogy (DbA) where inspiration is taken from the natural world to solve technical problems or challenges. Although Biologically-Inspired Design has been around for centuries (e.g. the flying machines of Leonardo da Vinci), nowadays, there is an increased interest for nature as a source of design inspiration. Three main motivations are often associated with Biologically-Inspired Design: (1) the proven performance of nature's solutions holds potential for avoiding a lengthy learning curve for solving technical problems, (2) bio-inspired design is frequently associated with identifying out-of-the-box solutions which is an appealing characteristic in an industrial context where continuous innovation is an important competitive weapon and (3) the potential for sustainable bio-inspired products addresses a general increase in environmental consciousness.Because the individuals tasked with developing novel concepts typically lack extensive biological knowledge, Biologically-Inspired Design requires overcoming a knowledge gap between the biological and technological domain. As there were (and are) no widespread tools or methodologies for bio-ideation in industry, it is generally assumed that spontaneous or accidental inspiration played an important role in triggering bio-inspired concepts. However, in an industrial context, a more pro-active approach is advisable. Therefore, in the last two decades, a new research domain emerged that aims at developing tools and methodologies to support Systematic Biologically-Inspired Design. At the onset of this doctoral study, a literature study exposed that the state-of-the-art in Systematic Biologically-Inspired Design struggles with scalability, meaning that only a fraction of human's knowledge about nature could be leveraged. The need for interactive work for each biological strategy during database expansion was the main cause. The identification of this common bottleneck led to the formulation of the first research goal, introducing scalability into the search phase of the Systematic Biologically-Inspired Design process. To meet this goal, this dissertation presents two approaches for Scalable Systematic Biologically-Inspired Design. First, an algorithm is proposed to scale one of the existing approaches for Systematic Biologically-Inspired Design (AskNature) by eliminating interactive classification into AskNature's Biomimicry Taxonomy. Second, a new scalable approach is proposed - named Scalable sEarch for systemAtic Biologically-InspiRed Design (SEABIRD) - that leverages both a patent and biological paper database to link products and patents (problems) to organisms and biological strategies (solutions).In order to meet the first research goal, a number of components, called supporting algorithms, are required that have applications which are not limited to the envisaged contributions. Hence, the second research goal is formulated as the development and dissemination of algorithms that enable the creation of tools to support Scalable Systematic Biologically-Inspired Design. The developed supporting algorithms are: webcrawling for biological strategy documents, mention and focus organism detection in biological strategy documents and multi-word product identification in patents. The value of these three supporting algorithms is demonstrated by their uptake in the two proposed approaches for Scalable Systematic Biologically-Inspired Design.status: publishe