1.1

Trends in publications and patents

While we have probably always been inspired by nature in our technological endeavours ranging from using sharp tools to imitate animals’ teeth and claws to Leonardo Da Vinci’s attempt to develop flying machines based on studies of bat and bird wing morphology, the systematization of this through the scientific disciplines of bionics, biomimetics and biologically inspired design first became established in the second part of the last century⁶. The exponential growth of research into biologically inspired design since then has been documented by a number of studies focusing on either the number of published papers^6–8 or the number of new patents^9,10.

In order to see if this trend is continuing we have updated the figures with current data. As can be seen in figure 2, the growth is continuing although there are some indications that it has been slowing down in recent years both in terms of total papers and in terms of the proportion of papers containing the words biomimetics, biomimicry, bionics and biologically inspired design. This could indicate that the growth phase is coming to an end and it will be interesting to see if the proportion of papers will reach an asymptote within the next decade. If so, the proportion of published papers would show a similar sigmoid pattern as that proposed for patents⁹.

Figure 2.

The number of published biomimetics papers in ISI Web of Science. The search was done on articles in the Web of Science core database using the search terms ‘biomim*’, ‘bionic*’ and ‘biologically inspired’ (TS=(biomim* OR bionic* OR biologically inspired)) in the title, keywords or abstract. To get proportional data the number of hits was divided by the total number of published papers (which were assumed to be found searching for ‘a’or ‘the’ in the title, keywords or abstracts - TS=(a OR the*). The bars indicate the number of published biomimetics papers, while the line indicates the proportion of these papers out of the total of all published papers.

Lepora et al. applied an information analysis method to determine the state of the art for biomimetics. They collected a dataset of about 18000 articles describing work on biomimetics and used them to answer four questions⁸. The first question looked upon where biomimetic research is published and they found that more than half the articles were published in 5 journals (Biomaterials, Bioinspiration and Biomimetics, Journal of Biomedical Materials Research A, Langmuir and Acta Biomateriala) while a single conference (ROBIO) had about a third of the papers. For the second question on how rapidly the subject of biomimetics was expanding it was found that the growth rate is increasing with around 6% per year which equals a doubling in 13 years. The third question on subjects within biomimetics revealed a long range of popular terms with ‘robot’ and ‘control’ being the most leading. The fourth question was if there are research communities within biomimetics and by looking at the connectedness of popular terms it was found that five distinct communities with recognisable themes could be identified (robotics and control, ethology-based robotics, biomimetic actuators, biomaterials science and structural bioengineering).

To update previously reported trends on the development of patent growth, the same search terms as used by Bonser¹⁰ were used. The program PatentInspiration¹¹ was used with the filter line: biomim* OR bionic* OR “bio* inspired”, searching in title, abstract and description and only granted patents, which resulted in a total number of 13,361 patents. The graph in Figure 3 shows a steady rise of patents, with some exceptional years. In 2002 there is a slight decrease, between 2007 and 2010 there are also two years (2008 and 2009) with significantly fewer granted patents, as also seen in 2012 and 2013.

Figure 3.

The number of granted patents over time illustrated via the program PatentInspiration. The search was done using the search terms ‘biomim*’, ‘bionic*’ and ‘bio* inspired’ (Filter=(biomim* OR bionic* OR “bio* inspired”)) in the title, abstract and description. The bars indicate the number of granted patents, while the line indicates the trend.

Bonser and Vincent¹⁰ found a peak of granted US patents (1666) in 2004, while our search only revealed 371 patents worldwide in 2004. This confirms that different database results are difficult to compare due to the individual parameters and search strategies¹⁰. Doing the same search in the European patent office search engine¹², in the title and abstract, a total of 6199 patents are found. The equivalent search in PatentInspiration found a total of 5786 patents. These results are closer, but it should be noted that a patent search has a different outcome than a literature search. While in literature, in addition to the title and the abstract, keywords are listed for readers to find papers easier, in patents, it is often the opposite. Even though patents are published for everyone to read, the writer of a patent tries to describe and mask the patent in a way that makes it very difficult to find. Therefore, it can be expected that the number of related patents is much higher than the number that is shown in the analysis.

3.1

Design model

Carrying out problem-driven BID requires that a design model is followed. The model specifies the sequence and content of the activities that lead the design team from a design problem to a design solution and could include the following activities: A. problem definition, B. problem framing, C. translation into biological search terms, D. search for biological analogies, E. selection and understanding of how the biological organism works, F. abstraction of key principles, G. design solutions, H. build prototypes and I. evaluate the quality of the developed solution.

The use of a design model can be illustrated with an example which have previously been described in SPIE²¹: How to make polymer needles. The problem definition (A) includes an analysis of the present use scenario and the improvement potentials. Today most medical needles for injections are made of metal to assure reliability and easy insertion. However, metal needles are considered dangerous waste after use, as it, unlike the rest of the syringe needs a dedicated waste stream. If the needles could be made of plastic, simpler waste handling would be possible for the combined syringe and needle. But plastic needles have much worse penetration properties than steel equivalents. The problem framing (B) is therefore how to ensure easy insertion of softer polymer needles. The biological search terms (C) could be ‘skin penetration’, ‘biting’ and ‘stinging’. The search (D) finds insects like the stinging mosquito (Figure 5 left) and the protection mechanism in cnidarians (Jellyfish), where venomous pointed cells are shot into the threating animal. Understanding the biological analogy (E) can be done through literature study, collaboration with a biologist or a detailed study of the biological organism. For the mosquito, a range of supporting mechanisms like skin stretching, vibration (Figure 5 right), impulse and serrations together explains why the fairly flexible and very thin stinging proboscis can penetrate skin. Each of these mechanisms can then be abstracted and described as key principles (F). The principles are used to propose design solutions (G) like a mechanical vibration mechanism that will lower the penetration force for a needle and hence reduce the requirements to its stiffness. This can be verified empirically with a prototype (H) and an evaluation (G) where the proposed solution is compared with similar solutions.

Figure 5.

The mosquito proboscis (left): Seen from above (a) and cross section (b). li: labium, l: labrum, h: hypopharynx, m: mandible, mx: maxilla. Vibration unit (right) mimicking elements of the mosquito stinging mechanism ²¹.

There exists a number of different design models covering all or some of the activities described above. The ISOstandard ¹ mentions a sequential model with 5 steps as shown in figure 6. The focus is on the search and understanding of biological analogies and on the abstraction from the biological model. The initial phases where the problem is analysed and formulated is not treated in the standard. The Biomimicry Institute applies a design model they call the design spiral shown in figure 7A²². It has 6 steps which cover all activities from the initial problem analysis (identify) to the final evaluation of the developed product. The shape of a spiral communicates that the model can be repeated in part or whole until a satisfactory result is achieved. Distinct contributions include the biomimicry taxonomy for problem framing, the Asknature database for search and life’s principles for sustainability evaluation. The design model used by the Georgia Tech BID group also comprises 6 steps as shown in figure 7B¹³. The steps in the model reminds of the steps in the design spiral. The major difference is the final evaluation step which is not directly mentioned in the Georgia Tech model. Distinct contributions are the 4 box and T-chart methods for problem analysis and comparison with biological analogy. The ParisTech group proposes a 9 step model shown in figure 8 which is inspired by the TRIZ methodology⁵. It covers the same activities as the design spiral, but subdivides some of the activities into separate actions. The DTU group uses a 5 step model shown in figure 9 which also covers all the steps from problem identification to final evaluation^18,23. The distinct contribution for this model compared to the others is the focus on the so-called biocards which are used to communicate the abstracted functional biological principles. The single steps in the 5 design models are compared in table 1 which shows how the models correspond to each other.

Figure 6.

Typical technology pull BID development process as described by ISO ¹

Figure 7.

A: The 6 step design model called the design spiral used by the Biomimicry Institute ²² B: The 6 steps in the Georgia Tech BID design model ¹³

Figure 8.

The steps in the Paris Tech problem driven biomimetic process related to a general problem solving process⁵

Figure 9.

The 5 step design model from the DTU biocard method ^18,23

Table 1.

Comparison of 5 design models for problem-driven BID. Light grey fields mark activities covered and darker grey fields are where the model has a distinct contribution.

3.2

BioTRIZ

For designers to successfully use biologically inspired design in a systematic way with the problem driven approach, we need efficient methods to translate biological abstractions into technical solutions. One promising way is to expand the Russian Theory of Inventive Problem Solving (TRIZ) to include biological solutions²⁴. TRIZ is a method for developing innovative solutions based on a large 39x39 matrix of contradicting features (for example wanting something larger, but also lighter). Each field in the matrix has a number of possible solutions (for example use of gasses or generation of lift) developed from hundreds of thousands of patents.

Julian Vincent and co-workers adapted and simplified this method into a 6 x 6 matrix (using the fields: substance, structure, space, time, energy and information) and populated it with both traditional engineering solutions condensed from the larger TRIZ matrix and with biological solutions analysed from about 2500 conflicts and their resolutions²⁵. While initially seen as a promising approach with some notable successes such as allowing a comparison between problem solving in engineering (TRIZ) and biology (BioTRIZ) showing only a 12% overlap²⁴, BioTRIZ has failed to gain widespread recognition or usage²⁵. However, the underlying assumption that TRIZ has significant potential for biomimetics remains and current efforts aim at utilising more of the original TRIZ tools than just the contradiction matrix including abstraction and transferring tools^26,27 and the development of a BioOntology that move away from the simple dialectic approach in the contradiction matrix and a one-to-one database to include the overlap of several inventive principles to better capture the complexity and dynamic nature of biological systems²⁵.

4.1

The cleaning effect of Lotus leaves

The Lotus effect is the mechanism by which small epicuticular wax crystals on the leaf surface of some plants, including the sacred lotus plant, allow the plants to self-clean by giving water a stronger adhesion to dirt particles than to the leaf surface²⁸. The mechanism was discovered accidentally when botanists, who were investigating the surfaces of leaves from a large number of species, observed that smooth leaves were much dirtier than rough leaves (step 1). From these initial observations, the exact nature of the roughness and its hydrophobic effect was quantified and the principles of their function extracted under the name of the Lotus effect (step 2 and 3). These steps were all completed by the botanists, lead by Willem Barthlott, who discovered the mechanism and they further suggested possibly applications in self-cleaning surfaces and paint (step 4 and 5)²⁸. Barthlott patented the Lotus effect in 1995 and the company Sto SE & Co acquired the patent and started producing the paint Lotusan (step 6 and 7), which is now commercially available and has demonstrated advantages including being more cost effective compared to traditionally paints²⁹. Thus, the Lotus effect is a clear example of a successful application of solution-driven BID.

4.2

The adhesion of gecko feet

Scientists and laypeople alike have long been marvelling at the extraordinary ability of geckos to adhere to, and climb on, vertical surfaces (step 1). However, the mechanism behind this achievement was first explained at the beginning of this century by biologists demonstrating that the thousands of hairs on each gecko foot each has more than 100s of minute spatula-shaped structures and that these adhere to surfaces via Van der Waals forces^30,31 (step 2). Within a few years, the Nobel prize winning physicists Andre Geim and colleagues had extracted the principles and produced a prototype ‘gecko tape’ made by microfabrication of dense arrays of flexible plastic pillars³² (steps 3, 4 and 5). However, challenges with durability and staying clean (dust accumulates which reduces adhesion – the gecko overcomes this by regularly cleaning its feet with its tongue) has meant that gecko tape is still not commercially available⁵ (step 7 not achieved). A further potential issue with gecko tape is the problem definition (step 6). While the main applications have focussed on static and semi-permanent adhesion (i.e. as a tape), the gecko uses the adhesion dynamically in a sequence of rapid attachments and detachments³³.

4.3

The material properties of spider silk

It has long been recognised that spider silk is a remarkable biomaterial allowing, among other things, spiders to catch fast flying and large prey with an ultralight weight web. It is particularly the frame threads (also used in the safety threads that spiders trail behind them when moving) that due to their combination of high strength and high elasticity has an exceptionally high toughness, which is necessary in order to absorb the high kinetic energy of the prey (step 1)^34,35. The biochemical structure of the silk is also fairly well understood and it is the composite nature of the silk consisting of amorphous flexible chains and stiff polyalanine arranged into hydrogen-bonded beta-pleated sheets (step 2 and partly 3)³⁶. However, the process of spinning occurring inside the spider, where the silk changes from a liquid phase in the glands and ducts to the solid phase extruded by the spinnerets via a complicated mechanical and chemical process is still not fully understood (step 3 not fully completed)³⁷. Thus, while biomimetic applications including in protective clothes and crash helmets have been identified (step 4 and 5), and the biochemical structure is known and the liquid silk proteins can be produced by genetically modified bacteria, the biomimetic spinning process has not yet been mastered and artificial silk is therefore currently much inferior to natural silk³³. However, while commercial products relying on spider silks’ unique material properties are a long way off (step 6 and 7 not achieved), the biocompatibility and organic breakdown of artificial or reconstituted silk makes it potentially very useful as suture threads in surgery and for other biomedical applications³⁸.

4.4

The defensive mechanisms of the bombardier beetle

The bombardier beetles (a subgroup of the ground beetles) display a fascinating and highly efficient defensive behaviour. They squirt hot noxious chemicals at would-be predators in a rapid and directed fashion (step 1). The exact biochemical method whereby this liquid is produced was already explained by biologists in the late 1960s³⁹. The liquid is produced when two separately stored chemicals (hydroquinone and hydrogen peroxide) are joined together in a chamber inside the beetle, where catalysts facilitate the decomposition of the hydrogen peroxide and the oxidation of the hydroquinone. This heats up the liquid close to the boiling point, which results in high pressure and an explosive reaction (step 2 and 3). Half a century later, engineers extracted the main principles by developing a numerical computational fluid dynamics model and using it to build a scaled up prototype of the combustion chamber with the aim of using the technique to improve gas turbines and produce pressure relief valve systems that can deliver a periodic pulsed spray (step 4 and 5)⁴⁰. The company Swedish Biomimetics 3000 is attempting to utilise this technology in µMIST® to develop the ability to spray highly viscous formulations, without requiring flammable and environmentally damaging propellants (step 6)⁴¹. However, to date no commercially available products have been developed (step 7 not fully achieved).

4.5

The fog collection by the Namib desert beetle

The Namib desert beetles are capable of extracting liquid water in the form of droplets from morning fog in the otherwise dry desert (step 1). The method with which this is achieved was explained by biologist Andrew Parker and engineer Chris Lawrence⁴². They found that the surface of the elytra (or wing casings) in these beetles is bumpy with a mixture of smooth hydrophilic non-waxy regions and textured hydrophobic waxy regions. Water droplets from surrounding fog forms on the hydrophobic peaks and rapidly grows in size until they roll down the elytra and into the beetle’s mouth as they adopt a curious headstand posture – also known as fog-basking behaviour. The elytra surface and posture can potentially be easily imitated by inclined textured artificial surfaces (step 2, 3 and 4). However, a later study found that several other beetle species with only smooth hydrophobic elytra were in fact more efficient fog collectors questioning the necessity for the textured surface for fog harvesting to work⁴³. Parker and Lawrence identified potential applications in the form of water-trapping tent and building covering or portable water-extractors (step 5). A prototype has recently been produced by depositing superhydrophobic nano-particles in a mesh placed on a hydrophilic copper sheet⁴⁴. The prototype was shown to be an efficient method of fog collection with excellent anti-corrosion ability (step 6 and 7).

4.6

Status of solution-driven BID

Thus, in conclusion, solution-driven BID has generated a large number of novel biomimetic applications, but as illustrated with the examples selected above, few have as of yet resulted in commercial successful products (only Velcro and Lotusan can be classified as successes). The main challenges are due to insufficient understanding of the biological mechanisms (spider silk), misapplying the process (dynamic attachment and reattachment in geckos used as inspiration for static attachment in tape) and/or problems with simplifying and replicating the process (spider silk and defensive mechanism in bombardier beetles). It is also worth noting that in all of the examples highlighted here, the first 3-4 steps (and for the Lotus effect, the first 6 steps) were completed by biologists before engineers and physical scientists took over illustrating the need for including both biologists and engineers in the solution-driven BID process.

5.1

Sustainability

Traditionally sustainability is described including three dimensions: the economic, the social and the environmental dimensions^2,50,51. United Nations has recently further concretized these dimensions into 17 sustainability goals⁵². The focus of sustainability driven BID is in many cases the environmental dimension, but especially the evaluation of the results from a sustainability driven BID-process is not meaningful without the economic dimension. In the following we will mainly be considering the environmental dimension of sustainability.

It is often assumed, that a BID product or system is automatically sustainable but that is not always the case. Speck and colleagues gives an example where a light weight product inspired by nature might save economic and environmental resources in building it, but the geometric shape might be so complex, that the production cost exceeds the saved material saving costs². Another example is poisonous animals and plants, which are the result of a natural production process but never the less are harmful to humans and other natural beings. The poisonous organisms serve their purpose within their ecological niche but the poisonous element cannot be taken out of context claiming sustainability. Speck et al. further describe how the descriptive, normative and emotional aspect of bio-derived developments are taken into consideration when looking at sustainability. The descriptive aspect aims at facilitating knowledge transfer between similar BID applications and proposes a classification into 17 classes for this purpose. The normative aspect aims at evaluating to what extend a proposed solution is sustainable while the emotional aspect deals with the comprehension of the product and the biological analogy. The Biomimicry Institute propose to use the Life’s Principles to evaluate the sustainability of a product – see the evaluation section later in the paper.

For the evaluation of environmental sustainability quantitative methods for life cycle assessment (LCA) are widely used. Examples are ‘pro Suite’, ‘PROSA’, ‘Calcas’ as described by Speck and colleagues². The LCA-tools calculate the impact of a product on the environment and translates it into indicators such as CO2-equivalents. The LCA-result is not an absolute measurement of sustainability but gives a relative number that can be used to compare new solutions to existing ones. To that end is normally defined a functional unit for comparison, for example the transport of one person a given distance in order to compare different transport solutions.

5.2

Biomimicry levels

The Biomimicry Institute defines 3 levels of biomimicry having increased complexity ^46,45. The first level of biomimicry concerns the mimicking of natural forms such as the imitation of frayed owl wing edges to achieve silent flight. The second level scopes the transfer of natural processes into technical solutions for example the way green chemistry is inspired by self-assembly of owl feathers at room temperature. The third level has the highest requirement as it covers the mimicking of natural ecosystems, also named as deep, holistic or eco biomimicry. It implicates the integration of all elements and the relationship between the elements in a more or less closed ecosystem. An example is the consideration of the owl feather as part of a forest and a sustaining biosphere. To achieve sustainable solutions level 3 biomimicry should be aimed for. Solutions found at the 2 lower levels need not being sustainable^46,53.

An engineering counterpart to the biomimicry levels is classification into a number of levels with increasing complexity^54,55. One is the distinction between the organism level, the behavior level, and the systems level, an approach, which stems from mechanical engineering design^56,49. We have used five levels to classify 5 examples of biomimicry: The system level on the top followed by products, components / modules, parts / elements and materials. Figure 10 place the example at the different levels and show their relation to the biomimicry levels. The technical examples seem to indicate the following tendency: Using natural forms are mainly to be found on the technical site at the element and component level. Natural processes (as second level) are implemented from the material up to products and devices, whereas mimicking ecosystems are mostly applied to the systems and product level.

Figure 10:

5 examples of biomimicry classified using the Biomimicry levels and a traditional engineering hierarchy of technical systems.

In megacities, there is increasing demand for climate regulation and purification of the air. This has increased the interst in natural mechanisms for air filtration and the increased plantation of trees and the concept of an urban green infrastructure⁵⁷. In extension to the contribution of parks and tree lined alleys, there is an increasing relevance of designing and implementing items of urban greenery to green roofs and facades⁶². However, as urban trees grow significantly faster than their rural conspecifics, they manifest a more rapid ageing and thus shortened lifetime which induces the need for new concepts integrating plants into megacities to ensure their contribution to the urban ecosystem⁵⁷. The example illustrate that it is not enough to mimic (or in this case copy) the natural organism but the ecological system in wood has to be considered.

One of the most well-known examples for buildings in tropical zones is the passive ventilation system at the Eastgate Building, Harare, Zimbabwe – which mimics the African termite mound ventilation system by the Architect Mick Pearce⁵⁸. The ventilation system reduces the energy consumption. It represents a smaller element in the building complex and illustrates that a termite mound being a more isolated system which utilize natural processes can be mimicked with positive sustainability consequences.

The efficiency of mechanical ventilation fans and energy producing windmills have been improved with inspiration from the bumps on humpback whale fins. The bumps reduces the drag and reduces the risk of stalling⁵⁹. The bumps are an example of how a natural form inspires to an improved component design.

Fire detection is traditionally done using energy consuming sensors that register either infrared radiation or smoke particles. Pyrophilous (fire-loving) insects utilize a mechanical sensor that detect very small volume fluctuations caused by infrared radiation⁶⁰. This can be mimicked to for passive, i.e. not consuming energy, sensors and hence be more sustainable. It is an example of a single element which inspires on the natural process level.

As earlier described does spider silk represent a unique combination of strength and elasticity^61,34,35. This makes it possible to produce more light weight impact resistant structures and thus save energy when transported. Even though the basic spider silk chemistry can be mimicked the unique properties require a better understanding of the spinning process, i.e. it is at the natural process biomimicry level.

7.1

Problem clarification

In any design process a crucial step is the clarification of the problem to be solved. If you do not ask the right questions the answers might be worthless. For the problem-driven paradigm this is the first initial activity where the engineering problem is analysed and the central challenging elements identified. For the solution-driven paradigm the clarification comes in at a much later stage in the development namely after the biological phenomena has been understood and mimicked and a suitable application is needed. For the solution-driven case the problem clarification is similar to the situation when new materials or technologies have been developed and a search for applications are initiated.

For problem-driven BID the clarification is often done by making a so-called design brief where background and context are described and characteristics of the desired solutions described. Since the design brief is intended to stimulate a creative generation of new ideas it is important that possible solutions are not mentioned in the brief since they would limit and constrain the possible solution space. The brief is normally a textual description but it can be complemented with a visual description where the nature of the problem is described, e.g. as a sequence of drawings describing the existing and desired future situation. A simple example could be the problem of dirty windows where a solution to cleaning is desired. A drawing of a dirty and a clean window would highlight the goal of the development work, but not describe the means to do the cleaning. Drawing for problem clarification in an integral part of the DTU biocard method²³. A major advantage of the visual representation is the communicative value for the development team and other stakeholders, where the focus can be kept on the core problem rather than on the details of the proposed solution.

Another tool is the 4-box analysis method proposed and used by the Georgia Tech BID-group⁶⁹. The 4-box analysis include a description of 4 elements: The operational environment, the functions, the specifications and performance criteria. Apart from enforcing a broader description of the problem the 4-box can be used in a multifaceted comparison of the problem and the proposed biological analogies using the so-called T-chart tool⁶⁹.

The Biomimicry Institute has developed a functional classification they call the Biomimetic taxonomy⁷⁰. The classification has 3 levels each with a number of categories. In total, there are 8 categories at the lowest level, 30 at the next level and 158 at the third. The use of the taxonomy can be illustrated with an example using the design of a cleaning mechanism for dry sanitation for developing countries⁷¹. The 3 level categories are ‘protect from physical harm’, ‘protect from non-living threats’ and ‘dirt/solids’. This classification can be used to search nature and for instance find how newly hatched small birds wrap the fecal matter as small parcels which the adult birds remove from the nest and in that way, protect it against predators. The bio-inspired solution was a wrapping mechanism that made encapsulation of fecal matter possible and hence removed the need for cleaning. The intention behind the biomimicry classification is to allow users to ask the question ‘how would nature solve this problem’ and make search easier. All the records in the Asknature database have been classified using the taxonomy making an effective search possible. However, the taxonomy is also very useful for searches in other databases or in literature.

7.2

Abstraction

For the first 3 BID paradigms, a key element is abstraction where central principles are identified and described. For problem-driven BID the abstraction takes place after the problem has been identified and described and the search for biological analogies is prepared. The problem is often described in context specific (technical) terms which will yield poor results if used for searching biological texts. For example, the technical problem in driving on icy surfaces can be formulated as ‘insufficient friction’ but a search on friction in biological text would not find all relevant biological analogies such as animal claws, slug slime or the gecko setae. An abstraction is therefore needed from the concrete technical formulation into more abstract terms (or even to more biology oriented terms).

In their natural search approach for BID, Shu and colleagues propose a set of techniques for abstraction and identification of relevant search terms to be used for the biological search ³. They strongly recommend to use verbs instead of nouns since the nouns are much more likely to find pre-conceived analogies. Verbs on the other hand will describe actions and hence be better at finding a greater variety of biological forms. As an example, they describe how the verb ‘protect’ will find a greater variety of phenomena than if the noun ‘cuticle’ was used. They also describe how certain verbs can be considered as biologically meaningful (significant or connotative) and if they occur within certain frequencies they can be considered as bridge words which are more likely to be helpful in the biological search.

Ahmed-Kristensen et al. gives further argument to why it is important to make abstractions^72,73. Using search terms well known to the person that searches causes a significant risk of design fixation – only solutions that are similar to what the person already knows about is found. They argue that truly novel ideas, i.e. ideas of concepts unknown to the person, are much more likely to occur when far analogies are used. Far analogies are much easier to find when searching nature compared to searching existing technical solutions provided that the search terms are not too closely related to concepts already known by the person.

7.3

Search

The search task has mainly been explored and developed for problem-driven BID but is also relevant for solution-driven BID where the challenge is to find relevant applications once the principles for an interesting biological phenomenon is understood. In the first case search is used to explore which biological phenomena that can solve a given problem. The search is typically done in literature available in public and commercial databases. Challenges in performing the search include minimizing search time and ensure that there are surprising and unknown search results. There are two generic types of search for BID:

To use natural language search, it is an advantage if the user has insight into biology and engineering. The important limitation of the natural language search are the different terminologies of engineers and biologists. The natural language search is less difficult if the field of interest within engineering and biology is narrow making it easier to learn the special terms and their translation. However, a narrow search also limits the amount of possible alternative solutions meaning that the search engineer is less likely to come up with radically new ideas. The use of structured tools as part of BID-tailored databases is a possible solution to this problem. BID-tailored databases with pre-defined categories and keywords have as prerequisite a mapping of attributes and functions of biological and technical systems. But as there is no consistent standardized and theoretical founded model of mapping engineering and biological matters, the available BID-databases differ widely.

BID-databases can be characterized by the level of concreteness. The more concrete content that is provided the more extensive the database and the user-interface. They also differ by the coverage of concrete biological solutions for engineering problems, and by the model for mapping biological and engineering items. Here we find

Vandemme and colleagues studied the impact of using databases to BID. The study was based on the AskNature database, and a control-group not using the tool and the overall conclusion was “The results consistently support a common claim used in favour of bio-inspired design, i.e. the expectation of identifying more out-of-the-box solutions.”⁸⁰. However, there are serious limitations for use of BID-databases. With the permanent increase in research results from biology, it seems an overwhelming assignment to keep updating the databases within an acceptable effort and time. This leads to the conception and application of tools for an automated classification “to integrate the large amounts of existing biological strategies much faster than currently is the case”⁸⁰. To overcome the obstacle of natural search and search by classification a range of methods are being developed including the method of using ‘association lists’, ‘bridging verbs’ and ‘BioScrabble’⁷⁹. Associative acting tools like the BioScrabble have been tested successfully with engineering master students, but despite positive results the application of tools like BioScrabble are not widespread since it mainly has been reported in German literature.

7.4

Understand

An important step in applying biological principles for designing products is the understanding of how the biological phenomena actually work. In many cases the functionality can be straight forward to understand for mechanically oriented engineers for example the biomechanics in walking, jumping, flying and swimming ⁸¹. But in general, it is easier for people with biological training to read and understand biological texts explaining the biological functional mechanisms. In particular, the more complicated causal relationships in biochemistry can be hard to grasp for the mechanically oriented engineers, even though the principles underlying the phenomena can be valuable for generating solution ideas.

The role of collaboration between engineers and biologist has been studied in Munich⁸². Different teams of either engineers, biologists or mixed teams were given short BID assignments. Even though the results were not statistically significant, there was a tendency towards better performance in the mixed teams. The cross disciplinary collaboration is an advantage not only for each discipline having access to extended competences but also for supporting the differences in terminology between the 2 fields. However, there are large differences in mind-set, values and working procedures between engineers and biologists which should be overcome for the collaboration to be successful.

The generic task of understanding is solved differently within the paradigms. Since much of the work on solution-driven BID is initiated and driven by biologists, understanding is typically a minor problem. Contrary for problem-driven BID which most often is initiated by engineers the task of understanding how the biological phenomena works can be a major challenge which limits the range of ideas that is explored.

7.5

Validation

To validate if proposed biological abstractions actually can be used to solve the initial problems physical or computer models normally needs to be produced. The models determine to what extend the functionality observed in nature can be simplified and mimicked. An example is the mechanism that prevent individuals in flocks colliding, e.g. in birds murmuration and in fish schools. The basic mechanism was explained by a movie animator who formulated 3 simple rules and showed that a computer program where the rules were implemented could simulate the flock behavior⁸³.

The validation can work as a motivating factor for cross disciplinary collaboration, and can to some extent explain the increased interest in biomimetics. The reason is the development within nano-technology and synthetic biology, where biologist find collaborators that can help them make good models and the engineers see opportunities to demonstrate how their expertise can be expanded to other areas. This can be seen within bio-photonics where physicists with insight into advanced computer modelling and thin layer manufacturing techniques makes it possible to understand and describe the advanced structural colour phenomena found in nature such as the bright blue colours in morpho butterflies and the antireflection in moth eyes⁸⁴.

7.6

Evaluation

Once a new product is developed using bio-inspiration it can be evaluated whether it solves the problem better than existing solutions. The immediate evaluation within a BID development project can be done based on the developed models looking at how well they satisfy the initial requirements. However, this evaluation only determines if the found solution is realistic but not whether it represents something new. To that end a set of metrics is being used to evaluate not only if the solution is (more) useful but also if it is novel. Shah⁸⁵ describes four metrics: Novelty, variety, quality and quantity. The degree of novelty is determined by searching for similar solutions. Quality tells us if the solution actually solves the problem in a satisfactory way. Quantity and variety tells how many alternative solutions have been proposed and how different they are. This is a measurement of the size of the explored solution space and serves as an indication of how likely it is that the optimal solution has been found. Another tool for evaluating the developed solution is the Indian SAPPhIRE model for causality^86,87. This model evaluates at which level of abstraction the proposed solution is different from existing solutions. The evaluation metrics can be used in problem-driven BID to judge the quality of the proposed solution and in solution-driven BID to assess if the biological principles for a given phenomenon have potential for further development.

The Biomimicry Institute uses the so-called ‘life’s principles’ in the Biomimicry DesignLens to evaluate the degree of sustainability a proposed solution represents⁸⁸. The principles can be used to measure proposed designs against natures sustainable benchmarks. There are 6 principles:

The principles are used in the evaluation process but not necessarily as an exact quantitative measurement. It is of course possible for some of the principles to compare exact values for the performance of a proposed design against existing solutions. But the main value lies in the evaluation process when looking on the performance with respect to each of the principles which can be more qualitative in nature, e.g. that recycling is possible and realistic.