Bilogically Inspired Design
Bilogically Inspired Design
A R T I C L E I N F O A B S T R A C T
Keywords: This paper reviews research on biologically inspired design, and has three main parts. The first part
Design
surveys examples relevant to three groupings of manufacturing research. The second part presents
Design method
general methods that support biomimetic design, including different approaches for the steps
Biologically inspired/biomimetic design
involved in identifying and applying relevant biological analogies for any given problem. The third
part details examples that illustrate the use of a general biomimetic design method, which identifies
analogies from natural-language biological information. Finally, insights and conclusions are drawn
and synthesized.
ß 2011 CIRP.
Fig. 4. Scanning electron micrographs of (left) a typical diatom frustule and (right)
nanometer scale silica biopatterning [57].
Reprinted with permission. Copyright Elsevier 2005.
2.2.1. Surfaces
Uniform silica coatings can enable desired optical, electrical,
chemical and mechanical properties for a multitude of applica-
tions. Pogula et al. initiated silica coating of glass using three
different types of biopolymers, and demonstrated that each alone
was able to catalyze the formation of silica films, and rapidly
produce a uniform coating, shown in Fig. 6 [70]. Li et al. reviewed
fabrication of biomimetic antireflective surfaces of silicon and
fused silica substrates [53].
Fig. 5. Biotemplating of sharkskin by Han et al. [35].
Interest in self-cleaning surfaces has motivated extensive
Adapted with permission. research into the superhydrophobic phenomenon observed on
biological surfaces. Most notably, the lotus effect has inspired
attributed to the diamond geometry of placoid scales on the several biomimetic surfaces and coatings, and refers to a
sharkskin surface, and to tiny 50-micrometer grooves on the hydrophobic self-cleaning property seen on lotus leaves. Although
surface of the scales. However, conventional micro forming a lotus flower tends to grow in muddy water, lotus leaves always
methods to replicate sharkskin have only been able to produce appear clean, and drops of water on the leaves appear crystal clear.
skins that reduce drag by up to 7%. Han et al. used actual sharkskin Two levels of surface structure on the lotus leaf, micro scale
in a bioreplicated mold for the production of synthetic sharkskin. mounds and nano scale hairs, shown in Fig. 7, cause water drops on
After preparing the sharkskin, they used hot embossing to the leaf to form high contact angles and minimal contact area with
replicate the external morphology including the surface of the the leaf surface [6,14,55]. Thus water and any dirt in the drop rolls
placoid scales containing the grooves. Sharkskin produced in this off the leaf easily, creating a superhydrophobic, self-cleaning
manner was shown to reduce drag by 8.25%. Compared to other surface. Mimicking the lotus leaf surface features, the lotus effect is
micro forming methods for sharkskin, this biomolding method commercialized in self-cleaning paints (Lotusan) and stain-
demonstrates high forming efficiency, good maneuverability, and repellant fabrics (NanoTex). Future applications could include
increased fidelity compared with live sharkskin, while being cost self-cleaning medical devices, cookware, building materials and a
effective. Potential applications include drag-reduced hulls for myriad of other products that benefit from self-cleaning or self-
ships and submarines. decontamination.
Biomimetic forming also includes fiber manufacture, e.g., A less well-known but also advantageous surface effect is
production of the spider silk that has inspired many with its observed in the dung beetle, which breaks up and compacts dung
strength and lightness. However, De Luca and Rey and Ellison into balls with neither dung nor dirt sticking to its body or legs.
et al. note that mechanical properties of spider silk are related to Tong et al. noted that the specialized geometry of the dung beetle’s
the spinning process in addition to the chemical composition of various body parts and geometrical embossing, shown in Fig. 8,
the silk [24,26]. Therefore, rather than relying on chemical or contribute to this surface effect [94].
artificial synthesis (which cannot produce silk of the same Potential applications of dung beetle surface geometry include
quality as spinning in vivo), efforts are underway to understand agricultural equipment, e.g., plows and tillers, and construction
and mimic the natural spider silk spinning process, which equipment, e.g., tunnel borers and excavators. After characterizing
involves low temperatures and water-based solvents. Saravanan the geometrical features of dung beetle surfaces, Tong et al. molded
notes that mechanical spinning may offer a more economical the dung beetle surface features on the surface of a furrow opener, a
method of production than even genetic recombination meth- type of plow [95], shown in Fig. 9. The biomimetic furrow opener
ods, where biological hosts such as hamsters and goats are demonstrated lower resistance and power requirements against soil.
implanted with silk genes to cause them to produce spider silk in Surface geometry rather than material can also reduce the wear
bulk [79]. on a surface exposed to abrasive substrates. Tong et al. studied and
Fig. 8. Details of female dung beetle C. ochus Motschulsky. a and b reprinted with
permission from Tong et al. [94]. Copyright Elsevier 2005. c reprinted with
permission from Tong et al. [95]. Copyright Elsevier 2009. (a) Female dung beetle C.
ochus Motschulsky, from Tong et al. [94]. (b) Cuticle surface details: (left)
stereoscopy image of clypeus and pronotum (right) scanning electron
microscopy image of pronotum convex domes [94]. (c) Ventral surface details:
stereoscopy photograph of: (left) ridges on longitudinal section of ventral surface
(right) ridges on ventral surface [95].
Fig. 12. Fish-based turbulence sensor, by Chagnaud et al. [10]. Reprinted with
permission. Copyright Society for Neuroscience 2008. (a) Top view of sensor with
pair of flexible micropillars spaced 5 mm apart in mean flow direction (left), pillar
length 500 mm (right). (b) Camera signal (left) and corresonding water velocity
vector field (right).
In another collaboration between the Technical University of ties; (2) environment with boundary conditions, physical and
Denmark and the University of Toronto, Shu et al. described a chemical properties; (3) assembly protocol or methods, e.g.,
method of microscrew ‘release’ based on abscission, which is repulsion and attraction; and (4) energy, e.g., internal or
involved in the release of fruit, petals, deciduous leaves in the transferred to the component through the environment [5].
autumn, etc. [86]. This case study will be discussed in detail as an Physical experiments involved placing magnetically encoded parts
application example of a generalized biomimetic design method in on a tray and assembling these parts into predetermined
Section 5.2. configurations through vibrations. These studies demonstrated
Lanzetta and Cutkosky exploit van der Waals forces to benefit feasibility of the self-assembly of mechanical parts, structures, and
assembly by fabricating biologically inspired hierarchical micro- ultimately of machinery and robots.
structures for synthetic dry adhesive applications [50]. Synthetic dry
adhesives may be superior to mechanical gripping and suction in 2.3.2. Modeling and controlling manufacturing systems
handling delicate materials such as glass, LCD panels and leather. Biological self-assembly and self-organization have also
To facilitate macroassembly, snap fits are ideal because they do inspired much research into production-level application of these
not require additional tools, and are generally time and cost concepts. As manufacturing environments become more demand-
efficient. However, challenges arise during disassembly for end-of- ing, e.g., shorter lead times, reduced inventory, increased number
life, including failure in the parts to be reused, for example, during of product lines and customization demands, researchers have
remanufacture. Detailed in another application example in Section aimed to better model, control and predict manufacturing systems
5.1, Hacco and Shu developed a repairable snap fit configuration with the aid of concepts from biology.
with predetermined break points, based on two biological Evolutionary and self-organizational biomimetic models have
phenomena, including plants replacing rather than repairing parts been developed to create more dynamic and adaptive manufactur-
such as leaves [33]. ing systems that better address increasing manufacturing com-
While snap fits facilitate assembly, during disassembly, they plexities. Ueda et al. have been a leading proponent of biomimetic
may present challenges in physical and visual access, or the need to modeling of manufacturing systems, developing the Biological
release multiple snap fits simultaneously. Willems et al. developed Manufacturing System (BMS) [98–101]. The concept of BMS can be
snap fits where a single operation activated through a pressure applied to the whole life cycle encompassed by a manufacturing
trigger can disassemble multiple parts simultaneously [115,116]. system, from planning to disposal. As in biology, the concepts of
One possible concept was inspired by a frog’s foot, which contains self-organization and evolution are central to the BMS model,
both rigid and flexible parts in one structure. This combination which is implemented through genetic algorithms, genetic
allows multiple tight locking points through the rigid parts, and programming, evolutionary strategy, cellular automatic, reinforce-
relative movement through the flexible parts during assembly and ment learning and immune networks. Also as in biology, two types
upon triggering the one-to-many disassembly operation. Two- and of information are found within the BMS, information passed from
three-dimensional finite element analyses showed that active generation to generation (DNA-type), and individually acquired
fasteners allowing for one-to-many disassembly achieved a 250% information based on learning (BN-type). In BMS, all entities such
increase in efficiency, enabling faster and more economical end-of- as work materials, tools, transporters, are autonomous organisms,
life disassembly. with products from raw materials expressing DNA-type informa-
‘Authorized’ disassembly may be desirable when a product tion, and manufacturing equipment creating the product through
contains relatively new microchips and memory components, or BN-type information. Self-organization within the BMS involves
large amounts of precious metals such as indium, platinum, and the use of attraction fields emanating from manufacturing cells to
gold. To discourage third-party reclamation activities, the OEM may attract jobs. DNA-type information would encode product type,
wish to render high-value components difficult to retrieve, such that while BN-type information would include machine selection that
only those authorized by the OEM can disassemble and reclaim could be based on past experience. Evolution with the BMS would
them, e.g., with a disassembly ‘key’. Saitou et al. developed a heat- involve DNA-type information directing work material to search
reversible snap-locator joint system inspired by enzyme-substrate for a work cell with a certain objective function, and occurs within
interaction [76]. This biological analogy was found by searching the the work material and work cell to meet the objective function.
biological literature with the term ‘lock and key’. Based on the Jin et al. described a model that mimics DNA-based capturing,
phenomenon that binding between the enzyme and substrate can be representing and applying design information to functions and
disabled by change in pH or temperature, Saitou et al. developed a changing operational situations, and cell-based differentiating of
locator-snap system in which strategic application of heat would function [44]. Jin et al.’s motivation is to enable large engineered
cause thermal expansion only in one plane, resulting in the release of systems to self-adapt to changes rather than to retool at a high cost.
parts. As a further deterrent to unauthorized disassembly, simulta- Tang et al. proposed a BioInspired Manufacturing System (BiMS) to
neous application of heat to specific areas could be required; with respond rapidly to production changes based on the ultra-short
the application of heat in the wrong areas destroying the part rather feedback loop of the neuro-endocrine immunity system [92].
than facilitating disassembly. Unlike the work of Ueda et al. and Jin et al., BiMS is not based on
Other research on assembly envisions machines and robots that DNA encoded information, but similar to the other two models, it
self-assemble in harsh or distant environments such as other incorporates the idea of autonomous cells to implement self-
planets. Bhalla et al. investigated self-assembling physical systems adaptive manufacturing control.
based on macro physical systems that only form under specific Another model borrows from concepts not only from evolution
conditions, e.g., sand dunes and galaxies [4,5]. Because these are but the study of evolution itself. To examine changes undergone by
physical systems rather than biological systems, they take a products and associated manufacturing systems, ElMaraghy et al.
bottom-up approach where information is encoded within the used cladistics to model product evolution as well as the associated
components and end environment of the system itself, rather than evolution of the manufacturing system [27]. Cladistics is a
a top-down approach such as using DNA to encode information. classification and modeling method that involves comparative
The benefit of a bottom-up approach is that no knowledge of the studies to find similarities among entities. A cladogram is
end result is required, mirroring the bottom-up self-assembly generated that hypothetically represents the evolutionary history
process found in natural physical processes. To understand natural of a group of entities, showing relationships and differences
self-assembly, Bhalla et al. performed both virtual and physical between the groups. The use of cladistics can provide a better way
modeling, and demonstrated that it is possible to create emergent of organizing entities as well as predicting the future behavior of
systems by mapping virtual self-assembly rules onto physical these entities. Further building on the use of evolutionary
systems. Bhalla and Bentley noted four requirements for artificial modeling of products and manufacturing systems, AlGeddawy
self-assembly: (1) components with shape, scale, material proper- and ElMaraghy hypothesized that like many species with a
680 L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693
common evolutionary course, products and associated production resolutions found in biology. The contradiction matrix for BioTRIZ
systems also co-evolve [1]. Establishing this link between products is only 12% similar to the original TRIZ, suggesting that different
and associated production systems may help prolong the life of principles are used to solve contradictions in biology. Notably,
manufacturing systems by predicting future behavior, thus biological solutions rely less on energy, and therefore the use of
facilitating decision-making. Tolio et al. reviewed and synthesized biologically inspired design is promising for the development of new
methods and tools that support the coordinated evolution (co- and sustainable engineering solutions.
evolution) of products, processes and production systems [93]. Research on methods that support biomimetic design in general
In the area of biomimetic production system control and falls under the two high-level categories:
scheduling, recent research includes genetic approaches to
scheduling by Chan and Chung [12] and pheromone-based 1. Methods to support search, retrieval and representation of
scheduling by Yu and Ram [122], Scholz-Reiter et al. [82], and biological phenomena for design.
Zhu et al. [125]. Scholz-Reiter et al. explain that, in an ant colony, 2. Studies to better understand and therefore support the
when given a choice of two paths between a food source and their application of biological analogies to design.
nest, initially half the ants will take the shorter path, and the other
half will take the longer path. During trips between the food source Clearly the two categories are related and how the first category
and the nest, ants deposit pheromones on the path, attracting other is implemented directly affects the second. Research under the
ants to the food source [81]. Since the pheromones on the longer above categories will be addressed in the order in which they are
path have more time to dissipate, over time, there is a weaker relevant during the biomimetic design process. Helms et al. [37]
concentration of pheromones on the longer path and a stronger identify two directions in biomimetic design:
concentration of pheromones on the shorter path, thus attracting
more ants to the shorter path. Many ant colony algorithms operate 1. Solution driven, where an interesting biological phenomenon
similarly to find the shortest scheduling route. inspires the search for potential applications, and
2. Problem driven, where a given problem motivates the search for
2.4. Conclusions for biomimetic manufacturing examples biological analogies that could help solve the problem.
This section presented a variety of biomimetic manufacturing The second direction, problem-driven is more widely practiced,
research in three groupings, with some overlap between groups 1 as well as promoted. Much of the general methods on biomimetic
and 2. Motivation common to much biomimetic research includes: design support this approach, and thus Section 4 is devoted to the
problem-driven approach. However, many commonly cited as well
1. Increased potential for sustainable manufacture, e.g., from as commercially available examples of biomimetic design, e.g.,
biomimetic forming processes in ambient environmental Velcro, are reported as solution driven. Therefore, the solution-
conditions to extending the life of manufacturing systems driven approach deserves some mention, and comprises the rest of
through biomimetic production systems modeling. the current section.
2. Ability to create ‘self’-enabled manufacture, e.g., from self- Finally, one of the recurring messages of this paper is that there
sharpening shredders to self-assembly of components to self- are several levels of biological organization, from the molecular,
organizing lineless production systems. e.g., DNA, to the ecosystem/biosphere that can be exploited for
biological analogies. Table 1, taken from Vakili and Shu’s initial
In many instances, more sustainable or self-assembling work [102] towards developing a natural-language approach to
manufacture provides alternatives to conventional methods. In identify biological analogies, lists these levels, as well as possible
some cases, biomimetic means provide a method where none was applications for these levels. Biological analogies are often recalled
easily accomplished or possible before, e.g., measurement of from the organ (e.g., lung, leaf) to organism (e.g., animal, plant)
turbulence and fabrication of highly structured porous materials. levels, supported by the many instances of existing biomimetic
Thus, biomimetic methods hold great potential for increasing design based on organs and organisms. However, biomimetic
manufacturing capability, sustainability and efficiency. However, design is more fully exploited by identifying analogies beyond the
not always available for the above examples are details on the obvious ones that come to mind, which motivates more objective
process of identifying and applying relevant biological analogies. search approaches.
Therefore, the remainder of this paper will be devoted to general
methods that support the identification and application of 3.1. Solution-driven approach
biological analogies to solve engineering problems.
Helms et al. identify the following steps observed in students
3. Methods that support biomimetic design using the solution-driven approach, where designers start with a
particular biological phenomenon in mind [37]. While many of the
Janine Benyus popularized the notion that humans emulate below steps are also involved in the problem-driven approach,
biological phenomena to design sustainable products and pro- definitions given here will be based on those by Helms et al.,
cesses [3]. A biological sciences writer, she founded the Biomimcry whereas the section on the problem-driven approach will contain
Institute, a clearinghouse for biomimcry researchers. A resource more perspectives on what constitute each step.
supported by the Biomimcry Institute is an online database of
nature’s solutions, available at AskNature.org. 1. Biological Solution Identification: Designers start with a biological
Julian Vincent, a biologist working in engineering, is developing phenomenon of interest in mind.
TRIZ (the Russian system for creative solution of problems) as the 2. Define Biological Solution: This step involves designers moving
main tool of biomimetics. Vincent et al. describe how TRIZ is adapted from structures and superficial mechanisms to a deeper
to support the transfer of knowledge from the biological to the understanding of the biological system. The example used
engineering domain in a method called ‘BioTRIZ’ [109,110]. In TRIZ, involves progressing from the initial understanding that abalone
problems are first expressed in terms of contradictions or conflicts. shell is hard, lightweight, resists impacts, into an understanding
Next, ‘inventive principles’ that typically solve those conflicts are of the complex interactions of composite materials that are
identified and applied. TRIZ was developed by analyzing over two responsible for this behavior. Functional decomposition typi-
million Russian patent certificates. Vincent et al. created BioTRIZ cally used for engineering problem definition can also assist in
using the same inductive approach as for TRIZ, but based on over 500 understanding the biological solution. In addition, Helms et al.
biological phenomena instead of patent knowledge. BioTRIZ define functional optimization as expressing functions in an
currently covers about 270 functions and 2500 conflicts and optimization problem or equation. As an example, the func-
L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693 681
Table 1
Levels of biological organization and possible applications, adapted from Vakili and Shu [102].
tional goals of moss, which are to reduce water loss and protect The majority of work in biomimetic design suggests that functions
reproductive structures from environmental stress, contradict are identified in this step. Functional modeling in engineering
the desire to increase surface area and position relative to the typically involves expressing the desired function to be fulfilled by a
sun for photosynthesis. Thus, the structure and placement of design solution. Complex high-level functions are decomposed into
moss optimize between these opposing sets of functional goals. simpler functions, any one, or combination of which can be solved
3. Principle Extraction: After the biological phenomenon is suffi- using biological analogies. A primary goal in functional description is
ciently understood, principles are extracted into a solution- to avoid limitation to specific physical solutions for as long as
neutral form, which involves removing reference to structural possible. Functions are typically represented by verbs. The use of
and environmental entities of the biological domain. For nouns, except as objects of functional verbs, is typically discouraged
example, instead of describing the abalone shell as ‘interactions as this may indicate that a physical solution is already in mind, and
between flexible proteins and hexagonal calcium carbonate not give other potential solutions fair consideration. However, the
deposits’, the principle is expressed as ‘tightly coupled use of adjectives also has potential, as adjectives may describe
composite material formation by alternating flexible and rigid desired qualities of the solution, e.g., flexible, transparent, etc.
structures for resisting impact.’ Although to some extent, the use of adjectives also assumes certain
4. Reframe Solution: Reframing the solution involves considering physical entities, e.g., surfaces that are flexible or transparent, versus
how humans would view the usefulness of the function the need to transmit light or allow for deflection.
achieved by the biological phenomenon. In an example given by Helms et al., students began with the
5. Problem Search: After reframing the biological phenomenon as problem of preventing shark attacks on surfers, and redefined their
usefulness to humans, human problems to which the principle desired function as camouflaging a surfboard. Here, camouflage is
can be applied are identified. the functional verb selected, and surfboard’, the object of the verb,
6. Problem Definition: An identified problem is defined using tools identifies the object to be camouflaged, and already limits the
such as functional decomposition and optimization. solutions to those applicable to surfboards, as well as specifies
7. Principle Application: The biological principle is translated into camouflage as a specific manner of protecting.
the engineering domain by introducing new constraints (and Other approaches may require redefining the original problem.
affordances), e.g., weight, flexibility, impact resistance and In TRIZ and BioTRIZ, the problem is first reformulated as a conflict.
manufacturing process criteria. Then, the principle is applied to Vincent et al. [110] give an example where tire chains are needed
develop a solution to the identified problem. on icy roads during winter to increase friction for safe driving.
However, chains will damage ice-free road surfaces, and conven-
tional alternatives such as changing between winter and summer
As could be expected, Helms et al. [37] observed a tendency in tires are inconvenient. Therefore, a tire that can instantly adapt to
students to fixate on the initial biological phenomena selected in road conditions is desired. The conflict becomes, ‘How can friction
solution-driven processes. This fixation is not ideal, as other between the tire and road increase, without increasing the weight
potentially more suitable biological phenomena may thus be (normal force) of the vehicle?’
overlooked. Helms et al. and others suggest as a step between problem
Lindemann and Gramann [54] also conducted a case study where definition and solution search, to reframe or ‘biologize’ the
they first studied a selected biological phenomenon, and then looked problem. This is done by redefining problems in biological terms,
for technical transfer opportunities. While they were able to often in the form of a question, e.g., ‘How do biological solutions
generate multiple concepts after studying the selected biological accomplish the desired function?’ The example given reframes or
phenomenon, they observed that selecting a ‘nice’ phenomenon in biologizes the function of ‘stopping a bullet,’ into considering how
the first place requires some knowledge of biology, and that biological entities ‘prevent, withstand and heal damage.’ However,
successful technical transfer may be an even more challenging task. this step must be undertaken with some caution, as it may
predispose the designer to specific solutions, e.g., biological
4. Problem-driven approach phenomena, and levels of biological organization.
The problem-driven approach requires that a problem be defined There are a number of ways to search for biological phenomena
sufficiently to enable a meaningful search for biological analogies. that are relevant as possible analogies to a given problem, each
682 L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693
with associated benefits and challenges. While the final goal is to based on the biological system. The second models a desired design
identify an analogy that provides a working solution, the initial based on customer needs, and looks for possible elements of
goal, ought be to identify a large variety of potential biological biological solutions from a repository [66].
analogies in an objective manner. The main disadvantage of searching through a database
Lindemann and Gramann propose a checklist of associations to specifically developed to support biomimetic design is that the
translate between technical functions and biology terms [54]. For search results are limited to what was entered into the database.
example, the technical functions of ‘change of the state of Depending on how the database is structured, bias may be
aggregation’ and ‘condense a gas’ are associated with biological imparted during the categorization of information as it is entered.
terms: nose passages, desert plants/animals, and leaves. A simple example is that while Velcro was developed from burrs,
Helms et al. [37] describe the following solution search should the biological entity of a burr be categorized under the
heuristics, relevant to several of the below methods as: engineering function of fastening? If so, other potential functions
or strategies that can be extracted from the burr may be lost.
1. Change Constraints – For example, generalize ‘keeping cool’, to
‘thermoregulation’. 4.2.3. Natural-language based search
2. Champion Adapters – Identify organisms or systems that survive In addition to building a database of biological phenomena
in extreme environments relevant to the problem, e.g., consider categorized for engineering use, another approach takes advantage
organisms living in desert or equatorial climates to find of the abundant biological information already available in
biological methods of ‘keeping cool’. natural-language format (e.g., texts, papers) by searching them
3. Variation within a Solution Family – Examine how different directly for relevant phenomena. This approach also avoids the
organisms solve a common problem, e.g., hearing, in different subjective and enormous task of cataloguing all of biological
ways, e.g., bats and echolocation, and how different solutions phenomena for engineering, and is summarized by Shu [87]. The
correlate with differences in the problem domain. natural-language approach is supported by Vandevenne et al., who
4. Multi-Functionality – Identify organisms or systems that solve propose a scalable approach for the integration of large knowledge
multiple problems simultaneously with single solutions. repositories in the biologically inspired design process [104].
While the natural-language approach overcomes the many
One limitation of the above is a bias towards ‘organisms’, as limitations of other approaches, there are also more complex
opposed to biological phenomena at any level, from the molecular challenges, as detailed below.
to the ecosystem. With or without incorporating the above, below
are approaches for finding biological analogies. 4.2.3.1. Natural-language sources. The initial source of natural-
language information is critical. This source should be written at a
4.2.1. Ask biologists directly level that can be easily understood by engineers with little or no
Certainly one obvious way of identifying potential biological background in biology. Life, the Science of Biology, by Purves et al.,
analogies for engineering problems is to ask biologists. AskNature a reference text for an introductory-level university biology course,
(http://www.asknature.org/) provides a social network of biolo- is an example of a suitable text [71]. In addition to ease of
gists specifically for this purpose. The advantage of simply asking comprehension, introductory-level texts tend to be general and
biologists is that the engineer, generally untrained in biology, does cover a wide range of organizational levels, from the molecular and
not have to search for and interpret the relevance and potential cellular to the ecosystem, such that potential solutions are not
application of information that may be unfamiliar. The disadvan- limited to one or two familiar levels.
tage of this approach is that one must have access to biologists. In Since a suitable search tool can be used to search any corpus, or
addition, biologists, as humans, may be biased towards their areas body of text, other texts can be substituted or added as required for
of expertise, rather than objectively recall and present a variety of the initial or subsequent search. The challenging task is the initial
phenomena as potential analogies. identification of relevant biological phenomena. Once relevant
phenomena are identified, further details can be found through
4.2.2. Search through database more advanced texts, research papers and traditional research
An obvious approach to address the limitations of asking methods. However, searching through more advanced sources
biologists directly is to attempt to capture their knowledge in a initially will generate results that are in more technical language
database. The advantages include more focused search results. and thus more difficult to understand. Due to the designer
Frequently, the same keywords used to categorize biological frustration this may cause, such results are more likely to be
phenomena, and often, past engineering solutions developed overlooked even if they are relevant. Therefore, initial results from
based on them, are presented as search keywords. Therefore, the a more basic text may be more effective for introducing the subject
‘relevance’ of information found is guaranteed since the same and confirming relevance to the design problem. An understanding
keywords used to categorize and enter information are used to and confirmed relevance of the subject then motivates further
then search the database. research and effort to understand the possibly complex details
Chakrabarti et al. developed a model to represent causality of needed to develop a solution.
natural and artificial systems, and used it to structure information
in a database of systems from both natural and artificial domains 4.2.3.2. Natural-language search keywords. As the same keywords
[11]. Bar-Cohen suggested the need for a documented database to used to categorize phenomena are often presented as possible
support biomimetic design [2]. Bruck et al. created a design search keywords, searching databases specifically created to
repository of bioinspired products and concepts to support support biomimetic design tends to be more straightforward than
bioinspired robotic projects for senior mechanical engineering searching natural-language sources. Unlike with databases, finding
students [9]. They also include a concurrent fabrication and relevant analogies in natural-language text is subject to the word
assembly technique to manufacture bioinspired designs. Wilson choices made by authors of the various natural-language sources.
et al. propose the use of reverse engineering and ontology to That is, multiple terms can describe and thus locate the same
structure a database to support biomimetic design [117,118]. concept.
Vattam et al. constructed a functionally indexed knowledge base of Vakili and Shu noted that synonyms are an obvious means to
structures and behaviors of biological and technological systems increase the number of matches for a given functional keyword
[106]. Nagel et al. used functional modeling and a design repository [102]. Chiu and Shu noted that troponyms, verbs that describe
to connect biological and engineering solutions, and suggest two specific manners of another verb, e.g., ‘ambling’ is a troponym of
approaches. The first functionally models a biological system, and ‘walking’ as it is a particular manner of walking, also represent
then seeks existing designs in a repository that can be improved suitable alternative keywords [19]. Keywords used to search for
L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693 683
‘break down’ enables ‘absorb’ in ‘Humans absorb amino acids by contain abstract entities, e.g., interest, equation, theory, etc., as
breaking down proteins from food’. the objects of search keywords are generally less applicable and
therefore not further considered.
Cheong et al. applied the above cases to identify biologically
meaningful keywords for function set verbs of the functional basis Ke et al. implemented the tagging of each word in a corpus with
[18]. Other translators were also asked to identify biologically metadata such as: its part of speech, and if a noun, whether it is an
meaningful keywords using the same cases as criteria, and abstract or physical noun, and possible senses. This information
confirmed the objectiveness and repeatability of the criteria. can be exploited to enable automatic hiding of matches that are
unlikely to be relevant [47].
4.3. Assessing biological analogies Mak and Shu observed that text descriptions of biological
phenomena that describe principles and behaviors in addition to
Another challenge in using biological analogies for design forms, tended to be more easily used by students as design stimuli
involves the human process of extracting relevant strategies from [60]. In addition, Cheong et al. noted that text descriptions of
the biological phenomena and applying these strategies to design biological phenomena that contain causal relationships, where one
problems. As explained above, the challenge of assessing biological action enables another action, are more likely to serve as useful
analogies for relevance to the engineering problem is most analogies for design problems [16,17]. Causal relations often
relevant to the natural-language search approach. Biologists explain how functions are achieved by behaviors. For example, in
pre-assess relevance before or while presenting potential biolo- the text excerpt, ‘Lysozyme is an enzyme that protects the animals
gical analogies. Relevance of biological phenomena is determined that produce it by destroying invading bacteria’ [71], the function
as they are entered into databases. of ‘protecting’ is, perhaps, counter intuitively enabled by the
In the natural-language search approach, the evaluation of behavior of ‘destroying’.
biological analogies affects the very selection of what is considered
relevant. The perception of relevance is very important, since, like 4.4. Applying biological analogies
with traditional search engines, search results may be presented as
a list of text excerpts. Mak and Shu observed that awareness of specific difficulties
that could occur when generating engineering concepts from
4.3.1. Categorization of search results biological analogies reduced their occurrence [62]. Therefore, this
A significant challenge in searching natural-language knowl- section presents categories of difficulties experienced by students
edge sources to identify biological analogies is the quantity and in detail rather than in summary.
quality of matches that may result. Depending on the keyword,
there can easily be an unmanageable quantity of matches. Below 4.4.1. Types of similarity
are strategies to address this challenge, while minimizing the loss Mak and Shu investigated how students applied descriptions of
of relevant information. biological phenomena functionally related to a given problem to
Words that frequently collocate with, or occur in the vicinity of, solve that problem [60]. Specifically, students were asked to develop
sought keywords can often summarize dominant biological concepts that result in ‘clean clothes’, by using the following
phenomena related to these keywords. For example, Chiu and description of a biological phenomenon by Purves et al. [71]:
Shu found that high frequency words, ‘predator’, ‘prey’ and ‘species’
for the keyword ‘eliminate’, describe how interactions between prey Barriers and local agents defend the body – skin is a primary
and predator species lead to one another’s elimination [19]. innate defense against invasion. The bacteria and fungi that
Ke et al. explored the use of Wikipedia categories to categorize normally live and reproduce in great numbers on our body
search results [46]. Also investigated were other means for surfaces without causing disease are referred to as normal flora.
categorization so that search results can be more quickly These natural occupants of our bodies compete with pathogens
summarized and skimmed. The most promising method is to sort for space and nutrients, so normal flora are a form of innate
by the subject acting on the keyword verb, and/or by the object defense.
that the keyword verb acts upon.
Four similarity types were observed between the description of
4.3.2. Evaluation and selection of biological analogies
the biological phenomena and the concepts developed using them
Evaluating the relevance of potential analogies across domains is
as stimuli. These similarity types are shown with examples in
a more complex process than evaluating the relevance of matches to
Fig. 14, whose axes are strategic accuracy and abstraction of
specific information sought. For example, in a traditional web search
biological entities. Details on the four similarity types are as
for ‘map of Paris’, it is far more obvious which matches actually
follows. In addition, related errors observed by Helms et al. [37] are
contain the map sought. However, with cross-domain analogies, the
identified under the corresponding categories, and indicated in
physical entities involved are likely to be different between analogy
Fig. 14.
and problem. Thus it is less straightforward to determine the
relevance, as well as usefulness of a given match. Described below
are challenges involved from the evaluation to the application of
matches, and known ways of addressing these challenges.
Hacco and Shu noted that matches with the following
characteristics could be eliminated right away [33]:
1. Literal implementation: A literal implementation involves using 4.4.2.1. Vaguely defined problems. Not unique to biomimetic design,
biological entities, e.g., bacteria, directly to solve the engineering problems that are vaguely defined, e.g., ‘lowering our dependence on
problem, e.g., by filling clothing pores to prevent dirt from oil,’ or ‘protecting a cell phone,’ are either too vague for functional
settling. Here, the biological entities are not abstracted, but modeling or do not limit the search space enough. Helms et al.
rather, used directly, with the same strategy between source and present the above examples improved as ‘more efficient allocation of
problem domains. Of relevance are the terms bioutilization and resources to reduce energy consumed in transportation’ and
bioassistance – the use of biological entities directly as ‘forming a scratch-resistant coating for cell phones.’
solutions. Bioutilization refers to the use of biological materials,
e.g., bamboo in floors, plant extracts in drugs, etc. Bioassistance 4.4.2.2. Poor problem-solution pairing. Designers were observed by
refers to using the actions of biological organisms to enable Helms et al. to match problems to biological solutions based on
solutions, e.g., plants to clean air, leeches to reduce blood superficial as opposed to functional similarity. The example given
coagulation, etc. involves matching ‘making a better dishwashing detergent’ with
This category is most related to the Helms et al. identified the ‘cleaning properties of the lotus leaf.’ While ‘cleaning’ is
behavior that designers use ‘off-the-shelf’ biological solutions to common between problem and solution, the lotus leaf self-cleans
‘do what it does’ instead of abstracting and applying the due to its surface structure, which typically cannot be changed by a
principles of the biological entity. The example given involves detergent.
using fireflies directly to emit light, instead of applying the Lindemann and Gramann also suggest an iterative approach to
principles that enable bioluminescence. bionics where failure to realize a solution based on a specific
2. Biological transfer: A biological transfer involves transferring the associated biological term requires abstracting the problem to a
biological entities, e.g., bacteria, into the solution domain, but different level, exploration of a different biological system, or both
without applying the strategy presented in the biological [54].
domain. For example, bacteria are used to provide the solution
of clean clothes by eating dirt. 4.4.2.3. Oversimplification of complex functions. Helms et al.
This category is most closely related to Helms et al.’s observed that designers tend to miss the significance of an
‘improper analogical transfer,’ where mechanisms important in underlying principle due to simplifying assumptions. The example
the inspiration source, but not relevant to the problem, were given involves using the term ‘simply writhing’, which over-
transferred to the problem. The example given describes how simplifies the deliberate, complex motion involved in writhing.
filters in a dog’s nose well suited for sorting through and
identifying several different scents are unneeded for a single 4.4.2.4. Simplification of optimization problems. Helms et al.
scent, but still transferred to the solution. observed that designers tend to fixate on a single biological
3. Anomaly: An anomalous solution involves neither the entities function instead of examining complex and competing biological
nor the strategy from the biological phenomenon. Some functions in forming optimization problems. The example given
anomalous concepts are due to lack of understanding. Other has designers expressing the structure of moss as optimizing
anomalous concepts are likely due to fixation on a few words in surface area to gather sunlight, and ignoring the requirements of
the text description while disregarding the overall strategy or the plant for protection and water conservation.
principle presented. Mak and Shu noted that the appearance in a Mak and Shu also noted that students tend to apply only one of
text excerpt of any word that is commonly used in engineering, several interacting strategies present in descriptions of biological
e.g., ‘motor’ in motor proteins, tends to result in concepts that phenomena [62].
involve motors, whether or not motors are relevant to the
phenomenon presented [61]. Although anomalous concepts can 4.4.2.5. Solution fixation. Designers were often observed by Helms
be novel as well as practical, they are not based on the biological et al. to fixate on the first biological phenomenon identified,
phenomenon, except perhaps as a source of random stimuli. exclude consideration of other phenomena, and prefer the initial
phenomenon to subsequent phenomena when comparative
The error Helms et al. observed that most closely relate is
evaluations were required. Only one out of nine teams was
‘misapplied analogy’, where matches were often forced into
observed to replace their initial biological phenomenon with
flawed solutions. The example given involves a two-way traffic
another. As biological phenomena were provided to students by
optimization algorithm derived from ant foraging behavior,
Mak and Shu [61,62], the fixations observed involved fixating on
resulting in an inaccurate model when applied to a throughput
particular solution modes as well as on irrelevant aspects of the
traffic optimization problem.
biological phenomena [61,62].
4. Analogy: The intended analogous solution accurately applies the
Santulli and Langella examined how students used biological
strategy from the biological phenomenon to the concept
analogies to design sports equipment. They observed that most
without transferring the biological entities, e.g., bacteria, into
projects were based on (1) animals, as opposed to plants, (2) simple
the solution. Gentner and Goel established foundations for
mental associations (e.g., jumping and frogs or grasshoppers, or
analogical reasoning relevant to this work [31,32]. More
light and fireflies), without necessarily a sound rationale, and (3)
recently, Vattam et al. developed a conceptual framework for
that formal imitation of structure was prevalent without
compound analogical design [105], studied the content of
necessarily matching the conditions between the engineering
different types of analogy in biomimetic design [106], and
and biological environment for which the structure is suited [78].
analyzed different kinds of analogies that occur at different
Wilson et al. also studied students using biological analogies as
stages in biomimetic design [107]. Sartori et al. implemented a
stimuli and report that exposure to biological examples increased
SAPPhIRE model to describe a biological system at various levels
design novelty without decreasing concept variety [119].
of scale and abstraction, and suggest that this specific
representation may support transfer at a particular level [80].
4.4.3. Analogical mapping tool
Yen and Weissburg support that the success of biologically
Although identifying the undesired categories of similarity
inspired design depends most critically on establishing an
reduced the quantity of misapplied biomimetic design, two
analogy at the appropriate level of abstraction [121].
persisting difficulties observed by Mak and Shu in students using
descriptions of biological phenomena were: inability to transfer
4.4.2. Additional errors observed in biomimetic design information from biology to engineering; and fixation on specific
In addition to the errors identified by Helms et al. already phrases of the description of biological phenomena [62]. This
categorized under the Mak and Shu’s [60] observations above, are motivated further studies where students were provided outlines
the following remaining errors [37]. of strategies to be applied to both biological and engineering
686 L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693
domains to facilitate analogical mapping. Compared to results resources needed to melt and reform components. To facilitate
without the mapping tool, the quality of generated concepts remanufacture, an identified design guideline is that product
improved. However, although participants who used the mapping features prone to failure should be made separable [83]. This
tool extracted strategies consistent with the biological phenomena separation enables the replacement of failed features and the reuse
presented, they continued to apply strategies to specific attributes of a component without labor- and capital-intensive repair. This
of the given problem. For example, participants tended to apply guideline however, contradicts design for assembly guidelines,
strategies to only one attribute of the problem domain, particularly since making failure-prone parts separate increases part count and
‘clothing’ in the clean-clothes problem, even though the resulting assembly cost. This apparent contradiction between ease of
concepts were not necessarily feasible. Because generated con- assembly and ease of remanufacture motivated the search for
cepts varied with selected attributes, a possible method to reduce potential biological analogies [102].
fixation and increase the variety of solutions developed, is to
instruct the designer to list the problem attributes, e.g., ‘clothing’, 5.1.2. Keyword and analogy identification
‘dirt’, ‘detergent’, and then develop ways in which the biological Not surprisingly, searching for the keyword ‘remanufacture’ did
strategies can be applied to each item in the list of attributes. not result in any matches in the biology text. Therefore, synonyms
Further work by Cheong and Shu and Cheong et al. support the and related words were required to find any matches. Other
idea that a template would be particularly helpful for descriptions keywords used include ‘repair’ and ‘correct.’
of biological phenomena that contain a causal relationship that
must be recognized and transferred across domains [16,17]. 5.1.2.1. Organ/organism level analogy. The keyword ‘repair’ located
a biological phenomenon at the organ to organism level in the
4.4.4. Functional models to aid comprehension Purves et al. text [71]:
Vakili et al. performed a preliminary study examining the role
of functional models in presenting biological phenomena as The defense systems of plants and animals differ. Animals
stimuli [103]. Study participants were asked to solve a micro- generally repair tissues that have been infected. Plants, on the
assembly problem using a set of biological representations, other hand, do not make repairs. Instead, they seal off and
including functional models, of leaf abscission for inspiration. sacrifice the damaged tissue so that the rest of the plant does
The visual aids provided to the designers were examined, and the not become infected. This approach works because most plants,
use of functional models of biological phenomena in particular was unlike most animals, can replace damaged parts by growing
critiqued. Solutions developed from the study were classified and new stems, leaves, and roots.
theories drawn on potential influences of the biological repre-
sentations. Observations, retrospective conversations with parti- The strategy derived from this phenomenon involves adding
cipants, and analogical reasoning classifications were used to new parts to replace damaged ones, rather than expending
determine positive qualities as well as areas for improvement in resources in repairing damaged parts, or sacrificing the entire
representation of the biological domain. Results suggest that organism. Applied to products, this strategy could involve planning
designers require an explicit list of all possible biological strategies. for parts to replace features that are likely to break, without
Nagel et al. explored the role of functional modeling to support repairing the broken feature or replacing the entire part that
biomimetic design [65]. Helms et al. studied how functional contained the feature.
representations of biological systems using the Structure-Beha-
vior-Function model facilitate understanding of biological systems. 5.1.2.2. Organ-system level analogy. A match found using the
They conclude that multiple representations, e.g., text, diagrams, and keyword ‘correct’, that describes fainting, follows [71]:
SBF models, should be provided to designers because each repre-
Blood must be returned from the veins to the heart so that
sentation is effective at facilitating the comprehension of different
types of biological knowledge, e.g., spatial, functional, causality, etc. circulation can continue. If the veins are above the level of the
heart, gravity helps blood flow, but below the level of the heart,
[38]. While functional models may help engineers understand
biological phenomena, Section 6 will submit that function may not be blood must be moved against the pull of gravity. If too much
blood remains in the veins, then too little blood returns to the
sufficient to identify the most suitable biological analogy.
heart, and thus too little blood is pumped to the brain; a person
may faint as a result. Fainting is self-correcting: A fainting
5. Examples
person falls, thereby moving out of the position in which gravity
caused blood to accumulate in the lower body.
Because the most detail is available for each, but especially for the
analogy identification, step of the biomimetic design process,
examples detailed in this section are from the natural-language The strategy derived from this analogy is to use preemptive,
approach. Most other published examples of biomimetic design tend defensive failure to prevent more serious failure.
not to specify how the biological analogies were identified. Case
studies using the natural-language approach include those in: design 5.1.3. Selected analogies and implementation
for remanufacture [33,85,102], authorized disassembly [76], micro- Fastening and joining are highly relevant to life cycle
assembly [84,86], sensing [52], redesign of fuel cells [22,47], protection considerations, as they clearly affect required assembly and
during hobbies [15], and protection from lunar regolith [23]. disassembly. Snap fits are often used as a fastening method due
Summarized below are three case studies executed in to their ease of assembly. However, snap fits frequently break and
chronological order over the span of nearly a decade, and research are difficult to repair during remanufacture.
results were applied to the case studies as they became available. Incorporating the preemptive failure strategy to the redesign of
Parts of the examples are presented in detail below to provide snap fits, Hacco and Shu specified predetermined breakpoints,
better context for more recent insights developed from them, shown in Fig. 15a, that may cause earlier failure, but the part
which are presented in Section 6. containing the snap fit feature can be more easily reused.
Incorporating the strategy to sacrifice and replace parts, a possible
5.1. Design for remanufacture planned replacement part is shown in Fig. 15b that can be used
once the sacrificial feature in Fig. 15a fails [33].
5.1.1. Introduction and problem definition
Remanufacture is a product end-of-life process that enables the 5.1.4. Conclusions for remanufacture example
reuse of product components. Over scrap-material recycling, This initial case study was performed to confirm that the
remanufacture offers advantages that include conservation of natural-language text search approach is able to identify relevant
L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693 687
5.2.3. Implementation
The abscission principle was applied abstractly to overcome
difficulties associated with ‘sticking’ as follows. The microobject is
released together with a part of the tool designated as sacrificial,
which can be of significant mass to take advantage of gravity. The
object can thus be easily released, and the sacrificial part of the tool
can then either remain with the microobject or be subsequently
removed.
For the specific application of inserting a 0.6 mm metallic
Fig. 15. Redesigned snap fit to facilitate repair by Hacco and Shu [33]. Reprinted
with permission. Copyright ASME 2002. (a) Snap fit redesigned with counter sink microscrew into a plastic counterpart, the abscission zone is
and break points. (b) Redesigned snap fit after failure and refurbishment. physically implemented as a polypropylene rod of 4 mm diameter
that is easily gripped and positioned by a small industrial robot
with six degrees of freedom and a specified repeatability of
and novel analogies to solve engineering problems. Analogies were 0.02 mm (see Fig. 16). The tip of the polypropylene rod is locally
found at several levels of biological organization, from the melted by heating, and then pressed over the head of the screw. The
molecular to the ecosystem, with only two detailed above. For contact with the screw results in solidification of the polypropylene,
both analogies, strategies were abstracted from the biological and a solid bond between the rod and the screw is formed. The robot
phenomena for implementation in engineering solutions. can now manipulate the screw into the plastic counterpart. Once the
screw is tightened into its final position, the resulting increased
5.2. Overcoming sticking in microassembly torque will break the bond between screw and rod.
Other concepts considered for the abscission principle included
5.2.1. Introduction and problem definition the use of ice or other intermediate material that could be
Size effects complicate the handling and assembly of micro- chemically dissolved. This however would introduce possible
mechanical parts. Specifically, surface-related forces, e.g., electro- contaminants that are clearly undesirable. The abscission principle
static, van der Waals and surface tension forces dominate is more broadly interpreted as a physically weaker zone between
gravitational forces. As a result, sticking between the micropart the plant (gripper) and the leaf (screw). Gravitational, torsional or
and the gripping device during release hinders the automation of other forces can be used to break this zone to separate the gripper
microassembly operations. and screw. In addition to relative ‘weakness’, a polymer was chosen
Shu et al. [86] describe the identification and use of biological as the intermediate-zone material due to its ability to form the
analogies to address the problem of ‘sticking’ during micro- very small geometric features that mate with and thus handle the
assembly [86]. Biological analogies were sought to develop microscrew.
concepts that enable the controlled removal of a microobject
from a gripping device. Release techniques based on DNA
transcription and abscission in plants led to concepts enabling
the automated handling of microobjects. Described below is a
concept implemented at the Technical University of Denmark.
5.3. Protection from lunar regolith ‘attach’, ‘protect’, ‘repel’, ‘push away’, ‘shield’, and ‘defend’. The
more promising biologically meaningful keywords were selected
5.3.1. Introduction and problem definition and used to identify the below matches.
In a collaborative project between the Canadian Space Agency,
Dalhousie University and the University of Toronto, biomimetic 1. Matches for ‘protect’: Most matches located with the keyword
concepts were developed to protect a Light Detection And Ranging ‘protect’ involve some physical protection and/or covering, e.g.,
(LIDAR) device from lunar regolith, or moon dust. Davidson et al. for embryos, seeds, etc. One match was not used entirely
report the use of biologically meaningful keywords to identify analogously, but did inspire a simple reconfiguration of the lens
biological analogies to generate solutions for protection required cap that would reduce the amount of regolith falling onto the
during lunar exploration [23]. lens during opening. This match involved a ‘clam’, inspiring a
In lunar exploration, regolith, or dust, is a significant problem lens cap configured as two halves of a bivalve that opens from
due to its pervasiveness, adherence, and abrasiveness, causing the top, such that regolith that has collected on top is less likely
premature failure of space suits, mechanisms, and scientific to fall onto the lens than from a single hinged cap, the current
instruments used on the moon. Unique characteristics of both configuration, shown in Fig. 17.
the lunar environment and lunar regolith prevent implementation 2. Matches for ‘repel’: The keyword ‘repel’ identified concepts that
of most obvious solutions. Therefore, biological analogies were use coatings, e.g., waxy coating on hair that repels water. One
sought to expand the range of possible and feasible solutions. match, involving the use of like charges to repel versus opposite
Two components of the lunar environment limit the choice of charges to attract, may be more conceptually novel than the use
materials. First, the thin lunar atmosphere approximates a hard of coatings, and led to the idea of using charge to manipulate
vacuum, and eliminates the specification of materials such as regolith away from LIDAR parts.
polymers, which will outgas in a vacuum and result in severe
physical degradation. Second, in the most common lunar 5.3.3. Selected analogies
exploration areas, temperatures range from 120 8C during the The biological analogies introduced above are detailed below.
day to 150 8C at night. Solutions therefore must accommodate An additional, more obvious biological analogy, not identified in
large and rapid temperature swings that occur during the change the manner described above, is also explored.
from day to night and vice versa, further limiting the selection and
arrangement of materials. 1. Bivalve class: Davidson et al. further investigated the bivalve
Both mechanical and electrostatic aspects contribute to lunar class of mollusks, which include scallops, clams, oysters, and
regolith adhering strongly to all surfaces. Without an atmosphere, mussels, and revealed the following relevant details. These
no wind rounds regolith particles. The sharp, jagged edges of organisms, evolved in particulate-laden environments, ‘protect’
regolith cause the mechanical aspect of adhesion, as well as the themselves from both predation and excessive particle intrusion
abrasion of mechanical seals. Furthermore, the small size of most with their shells. Bivalve shell geometry affords easy travel
regolith particles (below 70 mm), leads to infiltration of almost all through sand and mud, as well as protection from hazards. The
mechanical systems. Positive charging of particles by solar wind two shell halves meet at a 10–208 angle, which deflects sand
during the lunar day and negative charging by plasma electron away from mating surfaces, while the curvature of the shell
currents at night enable particles to cling to ungrounded sheds passing sand particles. Furthermore, the two halves of the
conductive surfaces and nonconductive surfaces. Finer regolith bivalve pivot to open and close using a flexible ligament, which
grains will levitate under this electrostatic charging and thus, dust is more resistant to particle fouling than standard mechanical
is present at the instrument level, even without mechanical hinges that require rotational sliding motion.
disturbance of regolith. 2. Charge: The strategy of using like versus unlike charges
The LIDAR device, shown in Fig. 17, is an optical instrument that identified by the keyword ‘repel’ led to the idea of moving
can detect particle concentrations kilometers above the instru- and keeping particles around the perimeter of the LIDAR
ment itself. A LIDAR consists of a high-powered laser that points receiver using high-voltage direct current (DC) electromagnetic
upward and an optical receiver. A cover, when closed, protects both fields. This method makes possible active dust protection with
the laser beam canister and the receiver lens. Improved protection the cover opened during operation, and exploits the electrostatic
is sought for the lens both during operation (when the cover is charge always present on lofted lunar dust that is not
open) and while idle. mechanically disturbed.
3. Human eye: Davidson et al. identified the human eye as an
5.3.2. Keyword and analogy identification obvious biological analogy for the protection of an optical
The engineering keyword ‘protect’ is identified directly from sensing system from dust. Tears not only hydrate the eye, but
the problem statement. In the translation from engineering to also provide a barrier between dust and the sensitive tissues of
biologically meaningful keywords, Cheong et al. grouped ‘protect’ the eye. The eyelid sweeps and actively transports foreign
with the functional basis words: ‘prevent’, inhibit’, and ‘shield’ particles to the perimeter of the eye, where they are flushed
[18]. Corresponding biologically meaningful keywords are: ‘cover’, from the eye with tears. However, directly implementing many
‘surround’, ‘inhibit’, ‘destroy’, ‘change shape’, ‘bind’, ‘release’, parts of the human eye analogy is not appropriate for this
L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693 689
problem. Use of fluids in the lunar environment is inappropriate 5.3.5. Conclusions for LIDAR protection example
because of the hard vacuum and large temperature ranges. The This case study supports a number of conclusions for
use of a mechanical wiper is not suitable because of the biologically inspired design. First, obvious biological analogies
abrasiveness of the particles and the sensitivity of the protected that come to mind, e.g., human eye protection, may not be the most
optical surfaces. However, suitable strategies include transport- suitable for a given problem. Second, biologically meaningful
ing particles to the perimeter of sensitive surfaces, and actively keywords, including those that are not obviously related to the
protecting these surfaces while in operation, which is accom- engineering function, e.g., ‘destroy’ (which was retrieved by
plished by the charge-inspired concept. protect), may lead to novel analogies and thus concepts. Finally,
although the clam analogy was not initially identified by a
5.3.4. Implementation functional link, the clam operates in a particle-laden environment
A dust protection system was prototyped based on the two similar to the lunar environment. Therefore the initial concept of
suitable analogies, and corresponding concepts are shown in two halves of a clamshell opening from the top was further
Fig. 18. The system is comprised of two protection mechanisms: a improved with additional characteristics, e.g., shell geometry and
Shape Memory Alloy (SMA) actuated, bivalve-inspired lid assem- ligament joint, of the clam that enable it to better cope with such
bly that protects the LIDAR device when not in use, and an integral an environment.
DC field generator in a circular collar that protects the LIDAR device
during use. 6. Summary, observations and conclusion
5.3.4.1. Bivalve inspired lid-assembly. Both the geometry and the This paper presented several specific instances of biomimetic
joint of the bivalve shell were incorporated into a two-piece cover design and manufacturing. However, not always available is how
system for the LIDAR device. The curvature of the lids facilitates the biological analogy was identified or selected. Therefore,
shedding of regolith and reduces regolith falling onto the lens area general methods, which support the identification and application
during opening. The bivalve ligament joint led to development of of biological analogies to any given problem, were described. One
an SMA hinge that also avoids the relative motion found in such general method, the natural-language approach to biomi-
traditional rotating joints, increasing resistance to particle fouling. metic design was illustrated through three examples involving a
Thin-film surface heaters supply heat to SMA actuators, which variety of applications. This section presents observations and
move each of the two lid sections. Davidson et al. provide further insights based on the specific instances of biomimetic manufac-
details of this implementation [23]. turing in Section 2, as well as the application examples in Section 5.
5.3.4.2. Particle management using charge. The receiver section of 6.1. Observations from biomimetic manufacturing examples
the base contains two semi-cylindrical insulated aluminum sheets,
which, oppositely charged, deflect charged particles away from the 6.1.1. Incorporate features on smooth surfaces
lens face. This solution works particularly well in the lunar A review of the examples presented in Section 2 reveal that
environment because of a weaker gravitational field compared to many biomimetic surfaces incorporate features that give them
earth. A charged particle that enters the electric field is drawn advantages over ‘smooth’ surfaces. Serrations on the mosquito
towards the oppositely charged plate, and is held at that plate upon proboscis and its bioinspired syringe needle reduce compression
contact. and nerve stimulation. Micro grooves on sharkskin scales and its
bioinspired surfaces reduce drag. Micro mounds and nano hairs on
5.3.4.3. Prototype test results. Davidson et al. report that the SMA lotus leafs and their bioinspired products enable them to stay
actuation system achieved more than 100 repeatable lid-opening clean. Geometrical embossing on surfaces of the dung beetle and
and closing cycles, and the high-voltage DC field controlled and its bioinspired furrow opener offer lower resistance and power
deflected 98% of incoming charged polystyrene particles, with an requirements. Ridges on scallops and the corresponding bioin-
average size of 1 mm, away from a representative surface. spired surfaces better resist abrasive wear.
While these phenomena were likely identified or recalled as
relevant based on observations at the organ to organism level, the
mechanisms that enable the phenomena of interest are generally
at much smaller scales. Therefore, there may be value in searching
directly for relevant phenomena at these scales, demonstrated
possible in the natural-language search approach to biomimetic
design. As engineering operates at increasingly smaller scales, not
only can these surface features be better observed, studied and
modeled, but they may also be better replicated in engineered
surfaces to confer similar advantages. This view is supported by
recent review papers on hierarchical features at the micro and
nano scales in natural and biomimetic surfaces, e.g., by Bhushan
and Jung for superhydrophobicity, self-cleaning, low adhesion, and
drag reduction [6], Sameoto and Menon for dry adhesion [77], as
well as Vincent [111] and Liu and Jiang [55]. Raibeck et al. point out
that while biologically inspired self-cleaning surfaces offer obvious
environmental advantages during use, environmental burdens
associated with current production techniques may offset these
advantages [72]. Therefore, there is likely potential for improving
techniques that produce biomimetic surfaces with respect to
environmental burden, perhaps by studying the corresponding
biological processes.
self-sharpening in rodent teeth and corresponding shredder activities needed in excision repair? Relegating these duties to a
blades. Gradients in attraction fields from manufacturing cells human repair technician confers no improvement over the current
attract jobs in the self-organizing Biological Manufacturing System situation in remanufacture. Further details on excision repair from
(BMS). Gradients in pheromone concentration allow ants and their a more advanced source mention conformation changes that result
bioinspired scheduling approaches to self-align to the shortest from the interaction of enzymes and contribute to the above
routes. The identification and exploitation of other possible activities in the repair sequence. The conformation-change
gradients can be used to realize other self-enabling activities in strategy was then matched to the engineering strategy where
engineering solutions. For example, Trask et al. propose biomi- failure induces a conformation change that helps release the part
metic self-healing polymer composites [97]. Furthermore, McKit- for replacement, e.g., self-removal during failure, somewhat
trick et al. studied mammalian structural materials including implemented already in the solution described in Section 5.
bones, teeth and tusks, horns and hooves for energy absorption Another biological analogy identified for the microassembly
mechanisms, and identify gradients in density as well as Young’s example also involves complex molecular interactions. This
modulus as possible biomimetic design strategies [63]. phenomenon was developed into a potentially feasible physical
concept, but was far more complex to implement than the
6.2. Observations from biomimetic method application examples polypropylene rods used to represent the abscission zone in the
solution described. Again, as engineering operates at increasingly
Below are retrospective observations from first-hand experi- smaller scales, it is possible that these molecular strategies can be
ence in the selection of biological analogies, extraction of biological better implemented directly.
strategies and their application in engineered systems, as In the lunar regolith example, the charge-based solution,
described in Section 5. although also ‘active’ in that energy is required, can be
implemented without the need for a physical actor, e.g., an
6.2.1. Consider analogies by environment in addition to function astronaut, a robot or a wiper to clear the lens of regolith while the
In addition to the search for and selection of biological analogies LIDAR is being used, all of which would be more complex as well as
based on functional similarity, as suggested by the majority of potentially unsafe.
biomimetic design researchers, the lunar regolith example also
highlights the role of environmental similarity. Specifically, while 6.2.3. Identify other preventive solutions in biology
the clam was identified more by superficial than functional Many of the biological analogies discussed in Section 5 are
characteristics, it turned out to provide a wealth of transferable based on preventive strategies in biology. For the remanufacture
strategies. The clam’s physical characteristics that are suited to its and LIDAR protection example, this is not surprising because of the
particulate-laden environment, e.g., shell shape that sheds sand nature of the problems and the keywords used for the searches. On
and ligament hinge less susceptible to particle fouling, are also well the other hand, it is interesting that a preventive strategy, i.e.,
suited to protection from lunar dust. Conversely, strategies from abscission in plants, was also identified and finally chosen to
the more obvious eye protection analogy, though more function- address the microassembly release problem.
ally similar, turned out to be much less directly transferable to the In biology, the consequence of failure or damage may be dire.
lunar environment. Therefore, strategies such as abscission and fainting are used
Helms et al. indirectly refer to the identification of biological before more critical damage to the organism occurs. There is likely
analogies by environment as well as function through champion an abundance of preventive strategies in biology that are highly
adapters – organisms that survive in the most extreme of the relevant to engineering, particularly the development of more
environment of interest [37]. Ke et al.’s use of adjectives in addition sustainable products and processes.
to verbs to search knowledge in natural-language text can be
applied to facilitate finding analogies by environment [47]. 6.3. Explore the applicability of ubiquitous biological phenomena to
engineering
6.2.2. Implement strategies passively to enable/replace active
solutions Many of the biomimetic manufacturing examples in Section 2
Although not detailed in this paper, earlier reports of both the were based on characteristics of organs or behaviors of specific
remanufacture and microassembly examples include biological organisms, e.g., mosquito proboscis serrations, self-sharpening
analogies that were not as obviously or successfully implemented rodent incisors, wood wasp reciprocal drilling, self-cleaning lotus
in the engineering solution. For example, identified using the leaf, drag-reducing sharkskin, wear-reducing ridges on scallops.
keyword ‘repair’ for the remanufacture problem are DNA repair However, in addition to the mosquito proboscis, serrations in
mechanisms. Specifically, excision repair targets damaged sections piranha teeth also confer similar advantages in cutting efficiency.
of a DNA molecule, including that which occurs during the life of In addition to the well-known lotus effect, dung beetles also have
the cell. Chemically damaged abnormal bases are excised and surface features that enable self-cleaning. Other biomimetic
replaced with functional bases. Clearly, this phenomenon is highly manufacturing examples are based on biological phenomena that
analogous to the repair of damaged parts during remanufacture. are common to many different organisms. For example, not only
The text on excision repair by Purves et al. follows [71]. does biomineralization enable diatom silica shells, but biominer-
alization in mammals and birds forms bone, and some bacteria
For example, in excision repair, certain enzymes ‘inspect’ the biomineralization involves iron and other metals to enable
cell’s DNA. When they find mispaired bases, chemically magnetic and gravity sensing, and iron storage. Therefore, it is
modified bases, or points at which one strand has more bases not surprising that biomineralization is used for two classes of
than the other (with the result that one or more bases of one applications (forming and coating) included in this paper alone.
strand form an unpaired loop), these enzymes cut the defective Hoeller et al. propose that capturing recurring solutions from
strand. Another enzyme cuts away the bases adjacent to and nature would support the design of sustainable products or
including the offending base, and DNA polymerase and DNA services [40]. Yen and Weissburg also support that the generality
ligase synthesize and seal up a new (usually correct) piece to and robustness of a particular biological strategy may be examined
replace the excised one. by how the strategy is implemented across many organisms [121].
An example of a ubiquitous biological phenomenon already
While the above description confirms relevance, it is difficult to successfully used in several fields of engineering is the optimiza-
apply the active strategies given. For instance, who or what can be tion method of genetic algorithms. Roy et al. note that while
used in place of enzymes in an engineering solution to ‘inspect, genetic algorithms represent the most popular technique for
find, cut, synthesize, seal up, and replace’ the defective parts, design optimization, swarm intelligence, also biologically inspired,
L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693 691
demonstrate potential for further application in design optimiza- [10] Chagnaud B, Brucker C, Hofmann M, Bleckmann H (2008) Measuring Flow
Velocity and Flow Direction by Spatial and Temporal Analysis of Flow
tion [75]. Fluctuations. The Journal of Neuroscience 28(17):4479–4487.
Towards the goal of identifying other ubiquitous biological [11] Chakrabarti A, Sarkar P, Leelavathamma B, Nataraju B (2005) A Functional
phenomena in a systematic manner, Cheong et al. are using Representation for Aiding Biomimetic and Artificial Inspiration of New Ideas.
Artificial Intelligence for Engineering Design Analysis and Manufacturing
natural-language processing to automatically identify strategies 19(2):113–132.
common to multiple biological units, e.g., biomolecules, organ- [12] Chan H, Chung S (2010) Genetic Approaches to Solve Economic Lot-Schedul-
isms, etc. This work may enable some common ground between ing Problem. Proc.6th CIRP-Sponsored Int. Conf. Digital Enterprise Technology,
Springer, Berlin, pp. 885–895.
the natural-language search and database approaches to support [13] Chen B, Liu Y, Chen S, Jiang S, Wu H (2008) A Biomimetic Spermatozoa
biomimetic design. Some phenomena have already appeared as Propulsion Method for Interventional Micro Robot. Journal of Bionic Engineer-
results to multiple natural-language searches for different ing Suppl.:106–112.
[14] Cheng Y, Rodak D, Wong C, Hayden C (2006) Effects of Micro- and Nano-
applications, e.g., abscission and plants sacrificing leaves for
Structures on the Self-Cleaning Behaviour of Lotus Leaves. Nanotechnology
remanufacture and microassembly. Such repeatedly relevant 17(5):1359–1362.
phenomena may warrant additional modeling to facilitate under- [15] Cheong HM, Shu LH, Stone R, McAdams D (2008) Translating Terms of the
standing, as well as be stored with past application examples. Functional Basis Into Biologically Meaningful Keywords. Proc. ASME Int.
Design Engineering Technical Conf., NYC, NY, August 3–6, DETC2008-49363
(DTM), .
6.4. Conclusion [16] Cheong HM, Shu LH (2009) Effective Analogical Transfer Using Biological
Descriptions Retrieved with Functional and Biologically Meaningful Key-
words. Proc. ASME Int. Design Engineering Technical Conf., San Diego, CA,
Biologically inspired design holds much potential for engineer- August 30–September 2, DETC2009-86680 (DTM), .
ing that is yet to be fully realized. Specifically, systematic access to [17] Cheong H, Chiu I, Shu LH (2010) Extraction and Transfer of Biological
biological information that is not limited to: 1. Obvious or chance Analogies for Creative Concept Generation. Proc. ASME Int. Design Engineering
Technical Conf., Montreal, Canada, August 15–18, DETC2010-29006 (DTM), .
analogies, 2. Well-known analogies already documented for [18] Cheong HM, Chiu I, Shu LH, Stone R, McAdams D (2011) Biologically Mean-
biomimetic design, and 3. Analogies at the organ-to-organism ingful Keywords for Functional Terms of the Functional Basis. Journal of
scale, is a first step towards better exploiting biological phenomena Mechanical Design 133:021007.
[19] Chiu I, Shu L (2004) Natural Language Analysis for Biomimetic Design. Proc. of
for engineering design. This paper described how the natural- 2004 ASME Design Engineering Technical Conf., Salt Lake City, UT, September
language search approach has been successful in identifying non- 28–October 2, DETC2004/DTM-57250, .
obvious analogies, from multiple levels of organization and scale, [20] Chiu I, Shu L (2005) Bridging Cross-Domain Terminology for Biomimetic
Design. Proc. ASME Int. Design Engineering Technical Conf., Long Beach, CA,
for a number of case studies. However, this approach also presents
September 24–28, DETC2005/DTM-84908, .
challenges that require further efforts. [21] Chiu I, Shu LH (2007) Biomimetic Design Through Natural Language Analysis
To best support the process of biomimetic design, it is vital to to Facilitate Cross-domain Information Retrieval, Artificial Intelligence for
objectively and systematically determine the needs of those using eng. Design. Analysis and Manufacturing 21(1):45–59.
[22] Currie J, Fung K, Mazza AG, Wallace JS, Shu LH (2009 July) A Comparison of
biomimetic design, including both novice and expert designers Biomimetic Design and TRIZ Applied to the Design of a Proton Exchange
from a variety of fields and backgrounds. Support tools should then Membrane Fuel Cell. 6th Int. Conf. on Innovation and Practices in Eng. Design &
address these needs, rather than be based on obvious solutions that Eng. Education, Hamilton, On, 27–29.
[23] Davidson M, Bligh D, Maloney N, McKnight C, Young W, Shu L, Potvin M-J,
already exist or approaches that first come to mind. Biologically Warkentin A (2009 July) Biomimetic Design of a Multi-Layered Dust Protec-
inspired design should not be limited to obvious biological tion System for Optical Instruments Operating in the Lunar Environment. 6th
phenomena and solutions already known to the designer, but a Int. Conf. on Innovation & Practices in Eng. Design & Eng. Education, Hamilton,
Ontario, 27–29.
systematic identification and application of biological analogies. [24] De Luca G, Rey D (2006) Biomimetics of Spider Silk Spinning Process. in
Similarly, research and tools that support the process of Brebbia’s CA, (Ed.) Design and Nature III: Comparing Design in Nature with
biomimetic design should not be limited to researchers’ personal Science and Engineering. WIT Press, Southampton, UK, pp. 127–136.
[25] Dong Q, Su H, Zhang C, Zhang D, Guo Q, Kiessling F (2008) Fabrication of
knowledge and experience, but aim to best address needs that are
Hierarchical ZnO Films with Interwoven Porous Conformations by a Bioin-
systematically identified from multiple perspectives of biologically spired Templating Technique. Chemical Engineering Journal 137:428–435.
inspired design. [26] Ellison M, Abbott A, Tuele F, Jeong S (1999) Biomimetic Manufacturing of
Fibers, NTC Annual Report, Nov 1999, M98–C07.
[27] ElMaraghy H, AlGeddawy T, Azab A (2008) Modelling Evolution in Manu-
Acknowledgments facturing: A Biological Analogy. CIRP Annals 57(1):467–472.
[29] Fraunhofer Institute for Environmental, Safety and Energy Technology
The authors are grateful for the helpful and critical review of (UMSICHT), 2007, Rodentics – Self-sharpening cutting tools, Annual Report,
pp. 60–61, accessed online April 23, 2011, http://www.umsicht.fraunho-
this manuscript by anonymous reviewers, as well as Michael fer.de/englisch/publikationen/.
Helms and Ashok Goel of the Georgia Institute of Technology. [30] Gao Y, Ellery A, Jaddou M, Vincent J, Eckersley S (2007) Planetary Micro-
Penetrator Concept Study with Biomimetic Drill and Sampler Design. IEEE
Transactions on Aerospace and Electronic Systems 43(3):875–885.
[31] Gentner D (1989) The Mechanisms of Analogical Learning, Similarity and
References Analogical Reasoning. Cambridge University Press.
[32] Goel A (1997) Design, Analogy, and Creativity. IEEE Expert 123:62–70.
[1] AlGeddawy T, ElMaraghy H (2010) Co-evolution Hypotheses and Model for [33] Hacco E, Shu L (2002) Biomimetic Concept Generation Applied to Design for
Manufacturing Planning. CIRP Annals 59(1):445–448. Remanufacture. Proc. ASME Design Engineering Technical Conf. and Computers
[2] Bar-Cohen Y (2006) Biomimetics – Using Nature to Inspire Human Innova- and Information in Engineering Conf., Montreal, September 29–October 2,
tion. Bioinspiration & Biomimetics 1(1):1–12. 2002, DETC2002/DFM-34177, .
[3] Benyus JM (1997) Biomimicry: Innovation Inspired by Nature. William Morrow [34] Hall SR (2009) Biotemplating: Complex Structures from Natural Materials.
& Co., New York. World Scientific Publishing Co., Pvt. Ltd., Singapore.
[4] Bhalla N, Bentley P, Jacob C (2007) Mapping Virtual Self-Assembly Rules to [35] Han X, Zhang D, Li X, Li Y (2008) Bio-Replicated Forming of the Biomimetic
Physical Systems. Proc. Conf. Unconventional Computing, Bristol, U.K., July 12– Drag-Reducing Surfaces in Large Area Based on Shark Skin. Chinese Science
14, . Bulletin 53(10):1587–1592.
[5] Bhalla N, Bentley P (2006) Working Towards Self-Assembling Robots at All [36] Harkness J (2001) A Lifetime of Connections – Otto Herbert Schmitt, 1913–
Scales. Proc. 3rd Int. Conf. Autonomous Robots and Agents, Palmerston, North 1998. Physics in Perspective In: http://www.thebakken.org/research/Schmitt/
New Zealand, December 12–14, . Otto.htm.
[6] Bhushan B, Jung Y (2011) Natural and Biomimetic Artificial Surfaces for [37] Helms M, Vattam S, Goel A (2009) Biologically Inspired Design: Process and
Superhydrophobicity Self-Cleaning, Low Adhesion, and Drag Reduction. Pro- Products. Design Studies 30(5):606–622.
gress in Materials Science 56(1):1–108. [38] Helms M, Vattam S, Goel A (2010) The Effects of Functional Modeling on
[7] Bobertag O, Feiat K, Fischer HW (1908) Über das Ausfrieren von Hydrosolen Ber Understanding Complex Biological Systems. Proc. ASME Int. Design Engineer-
Dtsch Chem Ges, 41. . 10.1002/cber.19080410353. pp. 3675–3679. ing Technical Conf., Montreal, Canada, August 15–18, DETC2010-28939, .
[8] Brott L, Naik R, Pikas D, Kirkpatrick S, Tomlin D, Whitlock P, Clarson S, Stone M [39] Hill J, Bart-Smith H, Barbier C, Humphrey J (2007) Investigation of a Bioin-
(2001) Ultrafast Holographic Nanopatterning of Biocatalytically Formed spired Whisker-Like Fluid Motion Sensor. ASME Int. Mechanical Engineering
Silica. Nature 413:291–293. Congress and Exposition, October 5–10, 2007, .
[9] Bruck H, Gershon A, Golden I, Gupta S, Gyger L, Magrab E (2007) Training [40] Hoeller N, Salustri F, DeLuca D, Pedersen Zari M, Love M, McKeag T, Stephens
Mechanical Engineering Students to Utilize Biological Inspiration During E, Reap J, Sopchak L (2007) Patterns from Nature. Proc. Conf & Expo. Society for
Product Development. Bioinspiration & Biomimetics 2(4):S198–S209. Experimental Mechanics, Springfield, MA, .
692 L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693
[41] Ikeshoji T (1993) The Interface Between Mosquitoes and Humans. University of [75] Roy R, Hinduja S, Teti R (2008) Recent Advances in Engineering Design
Tokyo Press, Tokyo, Japan. Optimisation: Challenges and Future Trends. CIRP Annals 57(2):697–715.
[42] Jaax K, Hannaford B (2004) Mechatronic Design of an Actuated Biomimetic [76] Saitou K, Shalaby M, Shu LH (2007) Bioanalogous Mechanical Joints for
Length and Velocity Sensor. IEEE Transactions on Robotics & Automation Authorized Disassembly. CIRP Annals 56(1):33–36.
20(3):390–398. [77] Sameoto D, Menon C (2010) Recent Advances in the Fabrication and Adhesion
[43] Junker S (1977) Ultrastructure of Tactile Papillae on Tendrils of Eccremo- Testing of Biomimetic Dry Adhesives. Smart Materials and Structures
carpus scaber R. et P. New Phytologist 78:607–610. 19(10):1–18.
[44] Jin Y, Zouein G, Lu S (2009) A Synthetic DNA Based Approach to Design of [78] Santulli C, Langella C (2010) Introducing Students to Bio-Inspiration and
Adaptive Systems. CIRP Annals 58(1):153–156. Biomimetic Design: A Workshop Experience. International Journal of Technol-
[45] Jones RW (2000) Near Net Shape Ceramics by Freeze Casting. Industrial ogy and Design Education .
Ceramics 20(2):117–120. [79] Saravanan D (2006) Spider Silk – Structure, Properties and Spinning. Journal
[46] Ke J, Wallace JS, Shu LH (2009) Supporting Biomimetic Design Through of Textile and Apparel Technology and Management 5(1):1–20.
Categorization of Natural-Language Keyword-Search Results. Proc. ASME [80] Sartori J, Pal U, Chakrabarti A (2010) A Methodology for Supporting ‘‘Trans-
Int. Design Engineering Technical Conf.& Computers and Information in Engi- fer’’ in Biomimetic Design. Journal on Artificial Intelligence for Engineering
neering Conf., San Diego, CA, USA, August 30–September 2, DETC2009-86681 Design Analysis and Manufacturing 24(4):483–505.
(DTM), . [81] Schmitt O (1969) Some Interesting and Useful Biomimetic Transforms. Third
[47] Ke J, Chiu I, Wallace JS, Shu LH (2010) Supporting Biomimetic Design by Int. Biophysics Congress, 297.
Embedding Metadata in Natural-Language Corpora. Proc. ASME Int. Design [82] Scholz-Reiter B, De Beer C, Freitag M, Jagalski T (2008) Bio-Inspired and
Engineering Technical Conf. & Computers and Information in Engineering Conf., Pheromone-Based Shop-Floor Control. International Journal of Computer Inte-
Montreal, Canada, August 15–18, DETC2010-29057 (DTM), . grated Manufacturing 21(2):201–205.
[48] Koh Y-H, Sun J-J, Kim H-E (2007) Freeze Casting of Porous Ni–YSZ Cermets. [83] Shu L, Flowers W (1999) Application of a Design-for-Remanufacture Frame-
Materials Letters 61(6):1283–1287. work to the Selection of Product Life-Cycle Fastening and Joining Methods.
[49] Kubota T, Nagaoka K, Tanaka S, Nakamura T (2007) Earth-Worm Typed International Journal of Robotics and Computer Integrated Manufacturing (Spe-
Drilling Robot for Subsurface Planetary Exploration. Proc. IEEE Int. Conf. on cial Issue Remanufacturing) 15(3):179–190.
Robotics and Biomimetics, Sanya, China, December 15–18, 2007, 1394–1399. [84] Shu LH, Lenau TA, Hansen HN, Alting L (2003) Biomimetics Applied to
[50] Lanzetta M, Cutkosky M (2008) Shape Deposition Manufacturing of Biologi- Centering in Microassembly. CIRP Annals 52(1):101–104.
cally Inspired Hierarchical Microstructures. CIRP Annals 57(1):231–234. [85] Shu LH (2004) Biomimetic Design for Remanufacture in the Context of Design
[51] Launey M, Munch E, Alsem DH, Saiz E, Tomsia AP, Ritchie RO (2010) A Novel for Assembly. Proceedings of I Mech E Part B: Journal of Engineering Manu-
Biomimetic Approach to the Design of High-Performance Ceramic–Metal facture, 218/3, 349–352. (Invited Paper).
Composites. Journal of The Royal Society Interface 7:741–753. [86] Shu L, Hansen H, Gegeckaite A, Moon J, Chan C (2006) Case Study in
[52] Lenau T, Cheong H, Shu L (2008) Sensing in Nature: Using Biomimetics for Biomimetic Design: Handling and Assembly of Microparts. ASME Interna-
Design of Sensors. Sensor Review 28(4):311–316. tional Design Engineering Technical Conferences & Computers and Information in
[53] Li Y, Zhang J, Yang B (2010) Antireflective Surfaces Based on Biomimetic Engineering Conference, September 10–13, Philadelphia, PA, DETC2006-
Nanopillared Arrays. Nano Today 5(2):117–127. 99398, .
[54] Lindemann U, Gramann J (2004) Engineering Design Using Biological Prin- [87] Shu L (2010) A Natural-Language Approach to Biomimetic Design. Special
ciples. Proc. Int. Design Conf., Dubrovnik 5, 18–21. Issue on Biologically Inspired Design of the Journal on Artificial Intelligence for
[55] Liu K, Jiang L (2011) Bio-Inspired Design of Multi Scale Structures for Function Engineering Design Analysis and Manufacturing 24(4):507–519.
Integration. Nano Today 6(2):155–175. [88] Sieber H (2005) Biomimetic Synthesis of Ceramics and Ceramic Composites.
[56] Livingston F, Sarney W, Niesz K, Ould-Ely T, Tao A, Morse D (2009) Bio- Materials Science and Engineering A 412:43–47.
Inspired Synthesis and Laser Processing of Nanostructured Barium Titanate [89] Stone RB, Wood KL (2000) Development of a Functional Basis for Design.
Thin-Films: Implications for Uncooled IR Sensor Development. Bio-Inspired/ Journal of Mechanical Design Transactions of the ASME 122:359–369.
Biomimetic Sensor Technologies and Applications, Proc. of SPIE, 7321, . [91] Tallon C, Moreno R, Nieto MI (2009) Shaping of Porous Alumina by Freeze
[57] Lopez P, Desclés J, Allen A, Bowler C (2005) Prospects in Diatom Research. Casting. Institute of Materials Minerals and Mining 108(5):307–313.
Current Opinion in Biotechnology 16(2):180–186. [92] Tang D, Wang L, Gu W, Yuan W, Tang D (2010) Modelling of Bio-Inspired
[58] Lottermoser A. (1908) Über das Ausfrieren von Hydrosolen Ber Dtsch Chem Ges, Manufacturing System. Proc.6th CIRP-Sponsored Int. Conf. on Digital Enterprise
41. . 10.1002/cber.19080410398. pp. 3976–3979. Technology, Springer, Berlin, pp. 1165–1174.
[59] Luz G, Mano J (2010) Mineralized Structures in Nature: Examples and [93] Tolio T, Ceglarek D, ElMaraghy H, Fischer A, Hu S, Laperrière L, Newman L,
Inspirations for the Design of New Composite Materials and Biomaterials. Váncza SJ (2010) SPECIES: Co-Evolution of Products Processes and Production
Composites Science and Technology 70(13):1777–1788. Systems. CIRP Annals 59(2):672–693.
[60] Mak T, Shu L (2004) Abstraction of Biological Analogies for Design. CIRP [94] Tong J, Sun J, Chen D, Zhang S (2005) Geometrical Features and Wettability of
Annals 53(1):117–120. Dung Beetles and Potential Biomimetic Engineering Applications in Tillage
[61] Mak T, Shu L (2004) Use of Biological Phenomena in Design by Analogy. Proc. Implements. Soil and Tillage Research 80(1):1–12.
ASME Design Engineering Technical Conferences and Computers and Information [95] Tong J, Moayad B, Ma. Y, Sun J, Chen D, Jia H, Ren L (2009) Effects of
in Engineering Conference, Salt Lake City, Utah, September 28–October 2, Biomimetic Surface Designs on Furrow Opener Performance. Journal of Bionic
2004, DETC2004/DTM-57303, . Engineering 63:280–289.
[62] Mak TW, Shu LH (2008) Using Descriptions of Biological Phenomena for Idea [96] Tong J, Mohammad M, Zhang J, Ma Y, Rong B, Chen D, Menon C (2010) DEM
Generation. Research in Engineering Design 19(1):21–28. Numerical Simulation of Abrasive Wear Characteristics of a Bioinspired
[63] McKittrick J, Chen P-Y, Tombolato L, Novitskaya E, Trim M, Hirata G, Olevsky Ridged Surface. Journal of Bionic Engineering 7(2):175–181.
E, Horstemeyer M, Meyers M (2010) Energy Absobent Natural Materials and [97] Trask R, Williams H, Bond I (2007) Self-Healing Polymer Composites:
Bioinspired Design Strategies: A Review. Materials Science and Engineering C Mimicking Nature to Enhance Performance. Bioinspiration & Biomimetics
30(3):331–342. 2(1):1–9.
[64] Meyers M, Lin A, Lin Y, Olevsky E, Georgalis S (2008) The Cutting Edge: Sharp [98] Ueda K, Vaario J, Ohkura KH (1997) Modelling of Biological Manufacturing
Biological Materials. Journal of the Minerals Metals and Materials Society Systems for Dynamic Reconfiguration. CIRP Annals 46(1):343–346.
60(3):19–24. [99] Ueda K, Vaario J, Fujii N (1998) Interactive Manufacturing: Human Aspects for
[65] Nagel RL, Midha PA, Tinsley A, Stone RB, McAdams DA, Shu LH (2008) Biological Manufacturing Systems. CIRP Annals 47(1):389–392.
Exploring the Use of Functional Models in Biomimetic Conceptual Design. [100] Ueda K, Hatono I, Fujii N, Vaario J (2000) Reinforcement Learning
ASME Journal of Mechanical Design 130:121102. Approaches to Biological Manufacturing Systems. CIRP Annals 49(1):343–
[66] Nagel J, Nagel R, Stone R, McAdams D (2010) Special Issue on Biologically 346.
Inspired Design of the Journal on Artificial Intelligence for Engineering [101] Ueda K, Hatono I, Fujii N, Vaario J (2001) Line-Less Production System Using
Design. Analysis and Manufacturing 24(4):521–535. Self-Organization: A Case Study for BMS. CIRP Annals – Manufacturing
[67] Nakamura T, Kato TO, Iwanaga T, Muranaka Y (2006) Development of a Technology (January 2001) 50(1):319–322.
Peristaltic Crawling Robot Based on Earthworm Locomotion. Journal of [102] Vakili V, Shu L (2001) Towards Biomimetic Concept Generation. Proc. ASME
Robotics and Mechatronics 18(3):299–300. Design Technical Conferences, Pittsburgh, PA, September 9–12, DETC2001/
[68] Oka K, Aoyagi S, Arai Y, Isono Y, Hashiguchi G, Fujita H (2002) Fabrication of a DTM-21715, .
Micro Needle for a Trace Blood Test. Sensors and Actuators A Physical 97– [103] Vakili V, Chiu I, Shu LH, McAdams D, Stone R (2007) Including Functional
98:478–485. Models of Biological Phenomena as Design Stimuli. Proc. ASME International
[69] Oh K, Chung J-H, Devasia S, Riley J (2009) Bio-Mimetic Silicone Cilia for Design Engineering Technical Conferences and Computers and Information in
Microfluidic Manipulation, The Royal Society of Chemistry. Lab on a Chip Engineering Conference, Las Vegas, NV, USA, September 4–7, 2007. DETC2007-
9:1561–1566. 35776 (DTM), .
[70] Pogula S, Patwardhan S, Perry C, Gillespie J, Yarlagadda S, Kiick K (2007) [104] Vandevenne D, Verhaegen P-A, Dewulf S, Duflou JR (2011) A Scalable
Continuous Silica Coatings on Glass Fibers Via Bioinspired Approaches. Approach for the Integration of Large Knowledge Repositories in the Biolo-
Langmuir 23(12):6677–6683. gically-Inspired Design Process. International Conference on Engineering
[71] Purves WK, Sadava D, Orians GH, Heller HC (2001) Life, The Science of Biology, Design, Technical University of Denmark, 15–18 August, .
6/e. Sinauer Associates, Sunderland, MA. [105] Vattam S, Helms M, Goel A (2008) Compound Analogical Design: Interaction
[72] Raibeck L, Reap J, Bras B (2009) Investigating Environmental Burdens and Between Problem Decomposition and Analogical Transfer in Biologically
Benefits of Biologically Inspired Self-Cleaning Surfaces. CIRP Journal of Man- Inspired Design. Proc. of the Third International Conference on Design Comput-
ufacturing Science and Technology 1(4):230–236. ing and Cognition, Atlanta, GA, .
[73] Ross BM, Wu LY, Lee LP (2009) Plasmonic Nanoflowers: Bioinspired Manip- [106] Vattam S, Helms M, Goel. A (2010) A Content Account of Creative Analogies in
ulation of Plasmonic Architectures Via Active Polymers, in Biomimetics and Biologically Inspired Design. AI for Engineering Design Analysis and Manufac-
Bioinspiration, edited by Raúl J Martı́n-Palma, AkhleshLakhtakia. Proc. of SPIE, turing (AIEDAM) Special Issue on Biologically Inspired Design Chakrabarti & Shu
7401, . (guest editors) 24(4):467–481.
L.H. Shu et al. / CIRP Annals - Manufacturing Technology 60 (2011) 673–693 693
[107] Vattam S, Wiltgen B, Helms M, Goel A, Yen J (2010) DANE: Fostering [117] Wilson JO, Rosen D (2007) Systematic Reverse Engineering of Biological
Creativity in and Through Biologically Inspired Design. Proc. 1st Int. Conf. Systems. Proceedings of ASME DETC/CIE, Las Vegas, DETC2007/DTM-
on Design Creativity, Kobe, Japan, 127–132. 35395, .
[109] Vincent J, Mann D (2002) Systematic Technology Transfer from Biology to [118] Wilson J, Chang P, Yim S, Rosen D (2009) Developing a Bio-Inspired Design
Engineering. Philosophical Trans of The Royal Society Physical Sciences Repository Using Ontologies. Proceedings of ASME DETC/CIE, San Diego, CA,
360:159–173. USA, DETC2009-87272, .
[110] Vincent J, Bogatyreva O, Bogatyrev N, Bowyer A, Pahl A (2006) Biomimetics: [119] Wilson J, Rosen D, Nelson B, Yen J (2010) The Effects of Biological Examples in
Its Practice and Theory. Journal of the Royal Society Interface 3(9):471–482. Idea Generation. Design Studies 31(2):169–186.
[111] Vincent J (2009) Biomimetics – A Review. Proc. IMechE Part H (J. Engineering in [120] WordNet 3.0 (n.d.). Accessed at http://wordnet.princeton.edu/.
Medicine), 223/8, 919–939. [121] Yen J, Weissburg M (2007) Perspectives on Biologically Inspired Design:
[112] Vizard F (2004) The Olympian’s New Clothes. Scientific American. 9 August Introduction to the Collected Contributions. Bioinspiration & Biomimetics
2004, accessed 28 April 2011, scientificamerican.com/article.cfm?id=the- 2(4). (Editorial).
olympians-new-clothes. [122] Yu X, Ram B (2006) Bio-Inspired Scheduling for Dynamic Job Shops With
[113] Wang J, Sato H, Xu C, Taya M (2009) Bioinspired Design of Tactile Sensors Flexible Routing and Sequence-Dependent Setups. International Journal of
Based on Flemion. Journal of Applied Physics 105(8):083515. Production Research 44(22):4793–4813.
[114] Wegst UG, Schecter M, Donius AE, Hunger PM (2010) Biomaterials by Freeze [123] Yu X, Tao J, Berilla J (2010) A Bio-Inspired Flow Sensor. in Varadan VK, (Ed.)
Casting. Phil Trans R Soc A 368:2099–2121. Proc. of SPIE, 7646.Nanosensors Biosensors and Info-Tech Sensors and Systems
[115] Willems B, Dewulf W, Duflou J (2006) Active Disassembly Fasteners and Their .
Potential to Enhance Environmental Impact Reduction. Proc. 16th CIRP Int. [125] Zhu Q, Gu Y, Zhang G, Zhang J, Chen X (2010) An Ant Colony Optimization
Design Seminar, Calgary, Canada July 16–19, 128–137. Model for Parallel Machine Scheduling with Human Resource Constraints.
[116] Willems B, Dewulf W, Duflou JR (2007) Active Snap-Fit Development Proc. 6th CIRP-Sponsored International Conference on Digital Enterprise Tech-
Using Topology Optimization. International Journal of Production Research nology, Springer, Berlin, 917–926.
45(18):4163–4187.