Res Eng Design (2009) 20:225–239
DOI 10.1007/s00163-009-0064-7
ORIGINAL PAPER
Affordance-based design methods for innovative design,
redesign and reverse engineering
Jonathan R. A. Maier Æ Georges M. Fadel
Received: 13 December 2008 / Accepted: 5 January 2009 / Published online: 12 March 2009

Springer-Verlag London Limited 2009
Abstract Rather than developing methods to address
problems as they occur, the effort in this paper is to formulate
methods based on an explicit theory. Methods developed in
this way have more scientific rigor because underlying
propositions and assumptions are clearly articulated, thus the
applicability and limitations of the methods are well defined.
The underlying theory used in this work is that of affordance-
based design, which has been developed by the authors in a
recent series of papers, and is based in turn on the theory of
affordances from perceptual psychology. This paper extends
affordance-based design into prescriptive methods. A broad
affordance-based design process is introduced together with
methods for documenting affordances, methods for design-
ing individual affordances, an affordance-based method for
reverse engineering and redesign, the affordance structure
matrix, and affordance-based selection matrices. Engineer-
ing examples used to illustrate the methods include the high
level design of an automobile, comfort to automobile
passengers, the meshing of gears, wear of gears, a vacuum
cleaner, and automotive window switches.
Keywords Affordances
Affordance-based design

Design methodology
1 Introduction
The existing state of the art in design science is character-
ized by much stronger methods than theory. Some methods,
J. R. A. Maier
G. M. Fadel (&)
Clemson Research in Engineering Design and Optimization
(CREDO) Laboratory, Department of Mechanical Engineering,
Clemson University, Clemson, SC 29634-0921, USA
e-mail: fgeorge@clemson.edu
such as those of axiomatic design are based on an explicit
theory (Suh 1990, 2001). Other methods, such as Quality
Function Deployment (Clausing 1994) are empirical in
nature, derived from what appears to work in practice,
rather than any underlying theory. Similarly, advances in
design-for-x methods are typically ad hoc in nature,
developed only in response to a particular need ‘‘for x’’.
Herbert Simon, in criticizing the lack of theoretical basis
behind design methods in the 1960s described such ad hoc
methods as ‘‘cook-booky’’ (Simon 1996). This is in con-
trast to the methods used in other engineering sciences, and
even mathematics, where methods to solve a particular
problem, say of heat transfer, are based on an underlying
theory, for example, Newton’s Law of Cooling. We pro-
pose that more rigorous methods in design should likewise
be based on theory first, and then validated and refined by
their use in industry.
The effort in this paper is therefore to present a series of
practical methods based on the theory of affordance-based
design elaborated in previous work by the authors (Maier
and Fadel 2001, 2002, 2003, 2005a, b, 2008). In Sect. 2,
the fundamentals of affordance-based design are summa-
rized and a high level method for affordance-based design
is presented, describing the steps needed to execute an
overall design project from inception to finished product.
Given the variability inherent in different design projects,
the steps described are of necessity somewhat general.
More specific methods for documenting the affordances
pertaining to a design project are offered in Sect. 3. A
method for designing individual affordances is given in
Sect. 4, and a method for reverse engineering and redesign
is presented in Sect. 5. A more detailed tool for reverse
engineering and concept exploration, the affordance
structure matrix, is discussed in Sect. 6, followed by
similar affordance-based selection matrices in Sect. 7.
123
226
Closing remarks and suggestions for future work are given
in Sect. 8.
2 A high level method for affordance-based design
In previous work, the authors (Maier and Fadel 2001, 2002,
2003, 2005a, b, 2008) discuss several theoretical aspects of
affordance-based design that can now be used to construct
prescriptive methods. Whereas the concept of affordance
was originally introduced in the field of perceptual psy-
chology (Gibson 1979), in the context of engineering
design, we define an affordance as a relationship between
two subsystems in which a potential behavior can occur
that would not be possible with either subsystem in isola-
tion (Maier and Fadel 2008). Such relationships are
particularly appropriate in describing the potential inter-
action between human users and artifacts, which are called
artifact–user affordances (AUA). However, affordances
can also describe potential behaviors between two artifact
subsystems, which are then termed artifact–artifact affor-
dances (AAA). An example of an AUA is a vehicle
affording transportation to its driver and occupants. An
example of an AAA is two gears affording power trans-
mission. In design, AAA usually exist at a lower level to
serve higher level AUA.
Important properties of affordances include:
• Complementarity: an affordance describes an interac-
tion between two subsystems. The affordance is relative
to both subsystems and cannot exist with respect to
either subsystem in isolation. For example, a single
gear cannot transmit power.
• Polarity: affordances can be either positive or negative,
depending upon whether the potential behavior has
beneficial or harmful consequences. A vehicle afford-
ing transportation to an occupant is a positive
affordance, whereas that vehicle affording injury to a
pedestrian is a negative affordance.
• Multiplicity: systems can have multiple affordances, as
in the vehicle example discussed above.
• Quality: affordances can be of varying quality depend-
ing upon how well the subsystems support the potential
behavior. The affordance of occupant comfort is of
higher quality in a luxury car, for example, than it is in
an economy car.
• Form dependence: affordances depend on the physical
structures of artifacts, unlike functions which are form
independent.
The design process can be viewed in terms of affor-
dances, in light of the concept of affordance, as discussed
previously by the authors (Maier and Fadel 2003; Maier
et al. 2008a, b). The impetus for any design project can be
123
Res Eng Design (2009) 20:225–239
understood in terms of creating and changing affordances.
The design process can be viewed simply as the specifi-
cation of an artifact that possesses certain desired
affordances, and does not possess certain undesired affor-
dances. An artifact with more (higher quality) positive
affordances is considered better, while an artifact with
more negative (or lower quality) affordances is considered
worse. Readers may consult the authors’ previous discus-
sion of these ideas for a more in-depth treatment.
Meanwhile, taking all these precepts into consideration, a
high level affordance-based design process is depicted in
Fig. 1.
The affordance-based design process begins with moti-
vation: a perceived market need, a novel idea, a scheduled
product redesign, etc. The parent company, if there is one,
would typically provide this information and initial product
cost targets, schedule targets, marketing targets, etc., to the
designer or design team.
In affordance-based design, the first task for the
designer(s), is to determine the AUA that the artifact
should have and not have. Because of polarity, the
designers should identify both positive affordances and
negative affordances (an aspect which is unique to affor-
dance-based design). And because of complementarity, the
affordances will depend on different users, so the design-
er(s) must identify the different users, grouping them as
convenient, and then interviewing various users to deter-
mine wanted and unwanted affordances. Following the
methods presented below in Sect. 3 for creating affordance
structures, the affordances should then be prioritized (the
highest priority affordance can also be considered the
design drivers); and finally, one or more affordance struc-
tures can be constructed. As described in Sect. 3, the
generic affordance structure template should also be con-
sulted in the process.
The second task for designers is to ideate to generate
concepts for the artifact’s overall architecture and com-
ponents. Various established ideation methods can be used
here, such as brainstorming, TRIZ (cf., Altshuller 1984,
1996, 1997, 2000), patent searches, Synectics (cf., Rourkes
1988), IDEO’s ‘‘deep dive’’ process (cf., Kelly and Littman
2001), etc. External references may also be consulted for
ideas, such as the Internet, traditional libraries, industry
catalogues, etc. Sketches of each concept are typically
produced using these ideation methods; however, these
sketches are particularly important for affordance-based
design, because of form dependence. Each concept should
be analyzed for how well it satisfies the desired positive
affordances with reference back to the affordances docu-
mented in the affordance structure(s) created in the
previous step.
The third task for designers is to analyze and refine the
affordances of the concepts generated in the previous stage.
Res Eng Design (2009) 20:225–239
Fig. 1 Overview of the
affordance-based design process
This involves modifying the characteristics of concepts in
order to modify their affordances, as well as analyzing the
negative affordances of each concept, and modifying their
characteristics accordingly to remove those negative
affordances. A helpful tool for this step of affordance-
based design is the Affordance Structure Matrix as
discussed in Sect. 6. Concepts from various architecture
concepts can also be combined, switched around, and
refined in order to modify the affordances of the overall
artifact. The construction of prototypes of concept archi-
tectures or components may also be needed in order to
better understand the affordances of each concept, again
because of form dependence.
The fourth task for designers is to select a preferred
architecture. Various selection methods can be used in this
process, including the Gallery method (Hellfritz 1978),
227
Pugh decision matrices (Pugh 1996, p. 167–176), a
Selection Decision Support Problem (Kuppuraju et al.
1985), Utility theory (cf., Hazelrigg 1996), or the affor-
dance selection matrices discussed in Sect. 7. However, the
decision should ultimately rest on how well each concept
satisfies the desired positive affordances while eliminating
or minimizing undesired negative affordances, giving
preference to concepts with higher quality affordances, and
preference to concepts with extra positive affordances.
Note that the theory of affordances does not suggest a
preferred selection method, but it does inform the criteria
to be used in the decision-making process.
The fifth task is to determine the artifact–artifact affor-
dances (AAA) that should exist between the subsystems in
the preferred architecture. For example, the transmission of
forces, heat, fluids, electricity, and information between
123
228
subsystems must be afforded. As these individual AAAs
are elucidated, they should be added to the affordance
structure(s) previously created.
The sixth task is to design individual affordances.
Accordingly, in Sect. 4, methods are described for
designing positive AUA, and positive AAA, as well as
methods for designing against negative AUA and negative
AAA. These methods must be executed on the preferred
concept architecture and components in order to specify
detailed artifact characteristics such that the detailed arti-
fact does indeed posses all the desired positive affordances
and does not possess any of (or at least minimizes)
the undesired negative affordances, as articulated in the
affordance structure.
3 A method for creating affordance structures
One of the first steps in popular function-based system-
atic design methods and their derivatives is the creation
of some type of function structure, a graphical repre-
sentation of required functions and some relationships
between them. After briefly reviewing the two most
prevalent types of function structures, a method is then
introduced for creating four different types of affordance
structures.
The purpose of a function structure is to organize the
functions of an artifact in a coherent manner. One type of
function structure called a function tree organizes functions
hierarchically, from a top-level general description of the
product function down to low level ‘‘atomic’’ functions that
can be decomposed no further. An example function tree is
shown in Fig. 2.
Another type of function structure organizes functions
in terms of flows; specifically, flows of energy, matter,
and information through and between product functions.
An example of this type of function structure is shown
in Fig. 3. This type of function structure can, but often
does not, capture the hierarchy that a function tree
captures.
Fig. 2 Example function tree (Shupe et al. 1987)
123
Res Eng Design (2009) 20:225–239
Fig. 3 Example function flow diagram (Pahl and Beitz 1996)
3.1 Affordance structures versus function structures
The purpose of an affordance structure is similar to that of
a function structure in that, whereas a function structure
organizes functions, an affordance structure organizes
affordances. However, because of the different properties
of functions and affordances, affordance structures cannot
be organized in the same ways as function structures.
Recall that a crucial difference between functions and
affordances is that functions are form independent, whereas
affordances are form dependent. The idea behind system-
atic function-based design methods [such as Pahl and Beitz
(1996)] is that customer demands and wishes can be
translated into functions, which can then be accomplished
by physical working principles, which are then embodied
by particular physical parts. It seems reasonable that fol-
lowing the rigorous steps of these methods should produce
more carefully considered designs; however, it is not clear
whether this is a consequence of the methods being func-
tional as opposed to merely systematic. In fact, there is at
least one noted example, coming from a very successful
design firm, that a systematic method, which does not
explicitly consider functions but which does move very
quickly to form solutions is creative, effective, and fast
(Kelly and Littman 2001). In this work, therefore, a sys-
tematic approach is maintained; however, the concept of
affordance is used as the fundamental theoretical construct,
rather than the concept of function.
3.2 A method to create specification level affordance
structures for original design
3.2.1 Step 1: understanding, gathering, and expressing
user needs in terms of affordances
The first step in building an affordance structure is to
express the users demands and wishes as affordances. From
a psychological point of view, this should not be difficult
since ecological psychology predicts that people perceive
their environment (including artifacts) already in terms of
Res Eng Design (2009) 20:225–239
Table 1 Affordances for a vehicle (annotations omitted)
Positive affordances: Negative affordances:
(The artifact must afford…) (The artifact must NOT afford…)
Transportation of occupants Injuring occupants
Transportation of cargo Injuring others
Comfort to occupants Damaging other vehicles
Entertainment of occupants Pollution to the environment
Communication to others
affordances (Gibson 1979). However, generally it is easier
to express affordances in terms of an affordance relation-
ship rather than a one word name for the affordance. For
example, it would usually be more appropriate to say that ‘a
vehicle must afford transportation’, than merely to list the
affordance of ‘transportability.’ Note that users include
not just end users, but everyone who will interact with
the artifact throughout its lifecycle, including designers,
manufacturers, marketers, salesmen, end users, mainte-
nance personnel, recyclers, trash haulers, etc.
At this point, the affordances should be grouped into
two categories: positive affordances (what the artifact
should afford) and negative affordances (what the artifact
should not afford). The output of this step is an unordered
list of annotated positive affordances and an unordered list
of annotated negative affordances. The annotations should
document from which users each affordance was suggested
and to which user each affordance pertains. As an example
artifact in this section, consider a typical passenger car. The
affordances of such a vehicle will be analyzed throughout
this section. However, the focus of this paper is on the
methods and not on any of the examples, per se, so the
vehicle example is pursued at a fairly high level. A short
list of positive and negative affordances for the vehicle is
shown in Table 1.
3.2.2 Step 2: apply generic affordance structure template
The second step is to compare the list of affordances with
the generic affordance structure template (Fig. 4). For
affordances that are present in the generic affordance
structure template, but do not show-up in the user affor-
dance list, the designer must decide whether those
affordances are indeed important (even though they did not
appear in the user affordances), or whether they can be
neglected for this particular project. The generic affordance
structure template thus serves two purposes. The first is to
guide designers as to what affordances in general they
should expect to provide. The second is to catch and
include affordances that might otherwise be missed. In the
passenger car example, there are many affordances in the
template that do not appear in Table 1. These additional
229
affordances pertaining, for example, to aesthetics, manu-
facturing, maintenance and the like, should be added to the
list of artifact affordances.
3.2.3 Step 3: prioritizing affordances
The third step toward building an affordance structure is to
prioritize the affordances. Note that, as a consequence of
Arrow’s theorem (cf., Franssen 2005; Scott and Antonsson
1999), this kind of ranking cannot objectively group user’s
preferences. Therefore, ultimately the prioritization of
affordances really only reflects the designer’s preferences,
but this should be informed, as much as possible, by user
information. Note that several affordances may share the
same priority level. The output of this step is a (tentatively)
complete list of prioritized annotated affordances. One
prioritization of the affordances for the passenger car is
shown in Table 2. As indicated in the high level design
procedure presented in Sect. 2, the resulting design
depends critically on the identification of pertinent affor-
dances, the subjective assignment of priorities to the
individual affordances, the understanding developed of
each affordance, and external factors such as time and
money available for the design project, market conditions,
legal regulations, etc.
3.2.4 Step 4: organizing the affordances into a structure
By priority The fourth step is to organize the list of
affordances into an ordered structure. This can be done in
several ways. One way to organize the affordance structure
is hierarchically by priority level. This organization is well-
suited to projects where most of the affordances are not those
shown in the generic affordance structure, and/or projects
for which there are many priority levels. This type of orga-
nization also leaves plenty of space to show annotations.
Often the most convenient way to represent this type of
organization is in the form of a table, as in Table 2 below.
By topic Another way to organize the affordances is by
topic, i.e., how the generic affordance structure is orga-
nized. This organization is well-suited to projects where
most of the affordances for the project are simply those in
the generic affordance structure, and/or projects for which
there are few meaningful priority levels. This type of
organization can be depicted graphically as a conventional
inverted tree, or as a sunshine type of tree as is the generic
affordance structure template. In either case, the priority
level for each affordance may be shown. However, one
drawback to this type of organization is that there is typi-
cally not enough space to include all the annotations unless
large format paper is used (which is sometimes done for
other organized graphs, such as House of Quality matrices).
123
230 Res Eng Design (2009) 20:225–239

Fig. 4 Generic affordance structure template

Table 2 Prioritized affordances for a ladder (annotations omitted) The annotations contain the necessary information to
Priority Positive affordances: Negative affordances: organize the affordances this way. A drawback to this
(The artifact must afford…) (The artifact must organization is that the same affordance can appear in
NOT afford…) multiple user groups. This can lead to a redundant structure
1 Transportation of occupants Injuring occupants if the user groups’ interests are relatively homogenous.
Transportation of cargo Injuring others
However, it is easy to indicate the priority level associated
Manufacture to workers
with each affordance. An example affordance structure of
2 Comfort to occupants Damaging other vehicles
this type is shown in Table 3.
Aesthetics to buyers and
occupants 3.2.5 Additional remarks
3 Communication to others Pollution to the
Improvement to owners and environment It is a relatively easy and quick process to draw an affor-
occupants Degradation of itself dance structure organized in any of these ways, so for any
Maintenance to owner and Frustration to occupants given design project it is good practice to prepare multiple
workers affordance structures to see which kind appears most useful
4 Entertainment of occupants for that particular project. Because the affordance struc-
Retirement to workers tures described above are created at the specification level,
Sustainability to the environment and because the specifications are usually given or inter-
preted as AUAs, a specification level affordance structure
typically contains mostly AUAs and relatively few AAAs.
An example of a topically organized affordance structure is
An affordance structure at this point, therefore, is neces-
shown in Fig. 5.
sarily incomplete until after a preferred architecture has
By user group Lastly, an affordance structure may also be been chosen, and the AAAs of that architecture have been
organized by user groups to which the affordances pertain. determined. Then those AAAs can be added to the original
This organization is well-suited to projects where different specification level affordance structure to complete a detail
groups of users have distinct and/or conflicting interests. level affordance structure.

123
Res Eng Design (2009) 20:225–239 231

Fig. 5 Topical affordance

structure for passenger car
example (annotations omitted).
White boxes are shared with the
generic affordance structure
template. Grey boxes are
specific to this example

Table 3 Affordance structure organized by user group (annotations In the case of AUA, we usually do not have control over
omitted) the properties of the user, but we do have control over the
User Positive Affordances: Negative Affordances: properties of the artifact under design. In the case of AAA,
group (The artifact must (The artifact must we usually have control over both artificial subsystems.
afford…) NOT afford…) Recalling the notion of polarity, which divides affor-
Occupants 1. Transportation 1. Injuring dances into those that are positive, i.e., beneficial, and
1. Transportation of their 3. Frustration those that are negative, i.e., harmful, then AUA can be
cargo positive (?) or negative (-), and AAA can be either
2. Aesthetics positive (?) or negative (-), this yields four major cate-
2. Comfort gories of affordances to design: ?AUA, -AUA, ?AAA,
3. Improvement and -AAA. The general method for designing any affor-
4. Entertainment dance is presented in Fig. 6.
Owner 3. Improvement 3. Degradation of their
3. Maintenance vehicle 4.2 An example of designing a positive AUA
Buyers 2. Aesthetics
Others 3. Communication 1. Injuring A positive AUA from the passenger car example is that the
4. Sustainability 2. Damaging their vehicles car must afford comfort to the driver and other occupants.
3. Pollution Comfort is an affordance that must be maintained across
Workers 1. Manufacturing many vehicle subsystems. For the purposes of illustration,
3. Maintenance consider only the comfort of the driver seat. Like any other
4. Retirement affordance, comfort has the property of varying quality.
Furthermore, because of complementarity as discussed
above, in order to create a desired affordance, in this case,
4 A method for designing individual affordances comfort, the designer must understand the particular
characteristics necessary on behalf of both the user and
4.1 About designing affordances artifact. Following the procedure in Fig. 6, we now need to
define and to understand what this affordance is. First, we
Since an affordance is a relationship between two inter- note that this is a positive affordance, something we want
acting subsystems such that a behavior is possible between the artifact to afford. Second, we note that the relevant
the two that is not possible with only one of the systems in users are drivers of the vehicle. These could be further
isolation, in order to design an affordance, we must con- divided into driver, front passenger, and rear passengers.
sider pertinent properties of the two different subsystems. The next steps involve information gathering. Comfort is

123
232
Fig. 6 Procedure for designing
individual affordances
determined fundamentally by the structure of the artifact,
taken with respect to the user. Fortunately, the physical
characteristics of seated automobile drivers have been well
studied, and are available in the form of anthropomorphic
data (cf., Webb Associates 1978). The challenge for the
designer is to match seat structure to human driver struc-
ture. Seat designers may also avail themselves of the many
existing examples of seat comfort already in existence,
from household and office furniture to existing vehicle
seats. The specific properties that determine comfort all
123
Res Eng Design (2009) 20:225–239
need to be identified. These result in this case in a list of
relevant anthropomorphic dimensions and seat dimensions.
The next step in designing the affordance entails iden-
tifying bounds and targets for each property identified
above. Since we are dealing with an AUA, we do not have
direct control over the properties of the user, but we do
have complete control over the properties of the artifact. A
principal goal of the design effort is determining the rele-
vant bounds and targets for the range of driver dimensions
deemed appropriate. A seat for a large SUV or truck needs
Res Eng Design (2009) 20:225–239
to accommodate larger size drivers than a compact car, for
example. This decision will impact the target height and
width of the seat back, elasticity of the seat components,
etc. Experimental studies are an invaluable tool in quanti-
fying such dimensions.
After bounds and ideal targets are identified for seat
comfort, the artifact’s affordance structure needs to be
consulted to identify which properties that impact the
affordance of comfort also impact other positive and neg-
ative affordances, for example, pertaining to strength of the
seat, or behavior in a crash. What is ideal for creating
comfort (such as a very soft seat) may not be strong enough
in critical loading conditions, for example. They key here is
to select property settings that maximize the affordance of
comfort subject to the constraints of maintaining previ-
ously designed affordances while attempting to safeguard
against negative affordances. As additional affordances
are designed, the designers must be vigilant to ensure
that previous affordances are indeed maintained, while
attempting to identify any new negative affordances. For
example, in the vehicle seat, a seat that is too comfortable
may in fact promote the driver sleeping, which is a
behavior to be designed against!
4.3 An example of designing against a negative AUA
Now that we understand how to create a positive AUA,
we need to consider the inverse problem, how to ensure
that a negative AUA is not present. Once again, this
involves understanding the affordance itself, and then the
characteristics of both the artifact and user that bear
upon that affordance. The goal in designing against a
negative affordance is to ensure that the afforded
behavior will not happen. Therefore, instead of specify-
ing artifact characteristics to lie within certain ranges,
and meeting certain targets, the designer must specify the
range that certain artifact characteristics cannot fall
within.
Recall that physical properties, i.e., the artifact’s char-
acteristics, are not affordances in and of themselves, but
they do determine what affordances exist with reference to
user characteristics. Therefore, the characteristics that
affect positive AUA are the same characteristics that affect
negative AUAs. The artifact, in other words, only has one
set of characteristics, and that is all the designer has control
over.
Therefore, when a designer identifies a range in which a
characteristic is not allowed to fall in order to avoid a
negative affordance, in reality all this does is constrain the
target bounds for that same characteristic with respect to
the positive affordances that are desired. These ideas are
best illustrated by another example. Consider the vehicle
seat from the previous section. As discussed above, a
233
negative AUA for the seat is conduciveness for sleep, or
‘sleepability’.
After defining the target affordance, the next step, as
before, is understanding that affordance. In particular, the
designer is interested in identifying what seat configura-
tions are especially conducive to sleep. Fortunately, again
the designer should utilize existing research into sleepiness
in vehicles (cf., Kanstrup and Lundin 2006). This research
will guide them toward identifying what ranges of seat
properties values are to be avoided, limiting, in effect,
design freedom as far as property settings for the earlier
designed affordance of comfort.
An important note here is that when desired positive
affordances and undesired negative affordances are iden-
tified early in the design process, then during conceptual
design, each concept can be judged relative to all identified
positive and negative affordances (see Sect. 8). Later,
specific affordances must be designed in order to maintain
the desired positive affordances while not creating any of
the undesired negative affordances. An affordance-based
view of the whole design process thus prevents the need to
change individual characteristics (such as seat properties in
the above example) later when it might be more difficult or
expensive to do so.
4.4 An example of designing a positive AAA
As in AUA, designing an AAA begins with defining and
understanding the affordance to be designed. Let us take as
our example in this section, the affordance of ‘‘turnability’’
of gears mentioned earlier. Recall that neither gear need
perceive that the other gear (or itself, or the combination of
the two) affords turning, since gears, being inanimate
objects (i.e., artifacts), cannot perceive anything. The gears
simply afford turning by virtue of their structure. Next, we
can consider an extrinsic purpose of this affordance (via the
designer), say, of power transmission.
The next step will be to identify what physical proper-
ties (i.e., structure) are necessary for both gears such that
when put together, the gear pair affords turning, and
thereby power transmission. As identified previously, for
the affordance of turnability to exist, at a minimum the two
gears must have compatible teeth, they must be positioned
in the same plane (say, for spur gears, not worm gears) and
close enough so that their respective teeth can mesh but not
bind, and the two gears must be able to rotate about their
respective center axes (Fig. 7).
Beyond those minimum requirements, the more detailed
information the designer has about the intended purpose of
the affordance/behavior, the more precise and detailed the
list of required physical properties can be. For example, if
the designer knows the magnitude of power to be trans-
ferred, the loads can be deduced and teeth sizes and profiles
123
234
Fig. 7 The affordance of
‘turnability’ for gears
designed accordingly, in addition to perhaps a lubrication
scheme. In any event, experimental studies or published
guidelines (cf., Litvin and Fuentes 2004) will be needed to
deduce appropriate targets for teeth design, gear spacing,
lubrication, etc.
Next, we must consider the effects of these considerations
on other affordances. Besides turnability, affordances of
gears include the ability to produce heat, the ability to pro-
duce noise, the ability to wear each-other, the ability to grind
other objects, etc. As these affordances are basically nega-
tive in nature, it behooves the designer to tackle each of these
negative affordances second, after first having set preferred
targets for the primary positive affordance, turnability. For
that the designer needs to know how to design against nega-
tive AAA, which leads us to the next section.
4.5 An example of designing against negative AAA
A negative AAA being present means that an undesired
behavior is possible. An example of such a negative
affordance, in the context of the gear example of the pre-
vious section, is the ability of the gears to wear each other.
Once again, the first step in designing (against) this affor-
dance is defining and understanding the affordance. It is
safe to assume that designers typically minimize wear so
that the gears will last longer, perform more efficiently,
generate less noise, etc. As in any AAA and as a conse-
quence of the property of complementarity, the affordance
does not exist with only one system; in other words, a gear
does not wear itself, but two gears wear each other.
Therefore, to design against the negative AAA of wear, the
designer must consider the characteristics of each gear that
lead to wear, which is the second step in designing an
affordance. For gears, three characteristics leading to wear
include the surface roughness of gear teeth, leading to
friction between gear teeth, how closely the gears are
spaced, and the geometry of the individual teeth.
Next, the designer needs to identify bounds and targets
for each of the above system properties. Wear can be
minimized (if not ever eliminated entirely) by changing
each of the artifact characteristics leading to it. For
example, polishing the gear teeth will decrease friction; the
123
Res Eng Design (2009) 20:225–239
Fig. 8 Optimizing the gear spacing parameter (Marlin 2004)
target for surface roughness would thus be zero roughness.
Friction will be further decreased by adding a lubricant. As
for spacing between gears, theoretically the optimum
spacing is such that the two gears’ pitch circles meet,
whereas the minimum spacing is where the addendum
circles meet, and maximum spacing is where the addendum
circles meet, see Fig. 8. Concerning the geometry of the
individual teeth, theoretically the involute profile is opti-
mum for spur gears, although design of experiments (DOE)
techniques could be used to look at variations in number of
teeth, tooth spacing, tooth thickness, etc., to determine the
best settings for each parameter in order to minimize wear.
The next step is to check to see if making any of these
changes in pursuit of eliminating a negative affordance
detracts from the positive affordances previously designed.
In this case, minimizing friction, and optimizing gear
spacing tend to improve the primary positive affordance of
turnability. However, changing the geometry of the indi-
vidual teeth affects the maximum load the gears carry, so
this trade-off between maximum load and wear should be
further studied before final property settings are chosen.
5 An affordance-based method for reverse engineering
and redesign
The various methods for creating function structures dis-
cussed in Sect. 3 are often (perhaps predominantly) applied
in reverse engineering and redesign scenarios, situations in
which a physical product already exists and is being ana-
lyzed to see how it works and could possibly be improved.
An existing product, however, can be understood in terms
of more than just the functions of its various parts. Just as
the function of a certain part can be ascertained by
inspection, so too can that part’s various affordances.
The information gleaned through a systematic evalua-
tion of the affordances of each part of the product can be
conglomerated into a complete detail level affordance
structure, including affordances between the product and
Res Eng Design (2009) 20:225–239
potential users (AUA) and affordances between the parts
themselves (AAA). A detail level affordance structure
produced in this way thereby mirrors the detail level
affordance structure that would have been produced by the
original designers of the artifact had they been using
affordance-based methods themselves. Even if the original
designers were not using affordance-based methods, say if
they were using functional methods or simply ad hoc
methods, the affordances of the artifact they produce still
exist and can be objectively studied with respect to certain
user groups.
After the creation of such a detailed affordance struc-
ture, the product has been effectively reverse engineered in
the sense that its operation should be well understood. The
next task toward redesign using an affordance-based
approach is to analyze the affordance structure for potential
improvements.
A detailed affordance structure can be analyzed for
improvement in the following areas:
For AUA:
• Identifying negative affordances that need to be
eliminated;
• Identifying potentially important positive affordances
that were left out;
• Analyzing the quality of existing affordances for
potential improvements in quality;
For AAA:
• Identifying parts with similar affordances that could be
standardized;
• Identifying redundant affordances, some of which
could be eliminated;
• Identifying groups of parts with related affordances that
could be conglomerated into modules.
6 Using the Affordance Structure Matrix for redesign
and concept exploration
As we saw in the previous section, the property of form
dependence results in a designer being able to analyze
concepts with respect to desired and undesired affordances
early in the design process, as well as within reverse
engineering. The Affordance Structure Matrix (ASM) is a
tool inspired by similar matrix based tools such as Design
Structure Matrices (DSMs) and House of Quality (HoQ)
matrices.
The Affordance Structure Matrix is a tool to compare
requirements information with physical structure at the
conceptual stage. Requirements information is interpreted
in terms of affordances following our previous work into
affordance-based design (Maier and Fadel 2008).
235
Specifically, requirements are organized into the four
categories of ?AUA, -AUA, ?AAA, and -AAA.
The interior of the ASM is populated by considering
whether each component of a product has a helpful (?),
harmful (-) or no () relationship to each affordance. The
‘‘roof’’ of the ASM is akin to a design structure matrix
(DSM) (cf., Steward 1981) that captures the relationships
between components. The left side of the ASM similarly
captures the relationships between affordances. For a
detailed description of the ASM and instructions for
populating an ASM, the reader is referred to our previous
work (Maier et al. 2007, 2008b)
An important difference between an ASM and other
similar matrices such as the House of Quality and design
structure matrices is the ability to capture whether rela-
tionships are helpful or harmful (±), not just existence (x)
or non-existence. Based on the identification of helpful and
harmful relationships, additional metrics are possible. In
particular, the total number of components that are helpful
with respect to each affordance can be calculated, as well
as the total number of components that are harmful with
respect to each affordance.
Similarly, the total number of affordances with which
each component has a helpful relationship can be calculated.
The total number of affordances with which each compo-
nent has a harmful relationship can also be calculated.
For each component, comparing the relative percentage
of helpful to harmful relationships gives a rough indication
of whether that component is doing more harm than good.
For each affordance, comparing the relative percentage of
helpful to harmful relationships gives a rough indication of
whether more components are helping or hurting to achieve
that affordance. For the overall product, the total relative
percentage of helpful to harmful relationships gives a
rough indication of how much room for improvement there
is; i.e., compared to an ideal situation where all compo-
nents are helpful.
A vacuum cleaner example is shown here for illustra-
tion purposes. A vacuum cleaner is chosen because it is a
frequently used consumer product studied in the literature.
Otto and Wood (2001) show a function structure for a
vacuum cleaner. Blackenfelt (2001) shows a strategic
DSM for a vacuum cleaner. An early version of an ASM
for a vacuum cleaner was presented by the present
authors (Maier et al. 2007; Maier 2008). A complete
ASM for a Eureka bagless upright vacuum cleaner is
shown in Fig. 9.
Turning our attention first to the bottom of the ASM in
Fig. 9, several parts, such as the power cord holder and
suction control knob, are seen which have 100 for the
‘‘Percentage Difference’’ in the bottom row. This indicates
that all of their relationships with the affordances are
helpful in nature. Note however, that the suction control
123
236
Fig. 9 Affordance Structure Matrix for a Eureka Bagless vacuum cleaner
knob is related to three affordances, whereas the power
cord holder is related to only one affordance, indicating
that the suction control knob may be more critical or more
sensitive to change than the power cord holder. Other
components have negative percentage differences (marked
with red backgrounds), such as the belt, with -33 per-
centage difference. A negative percentage difference for a
component indicates a potentially troublesome component
that is an opportunity for improvement.
Looking now at the right side of the ASM in Fig. 9,
again several affordances (such as dirt removability) have
100 percentage difference, which indicates that all the
components related to those affordances are helpful.
However, again several affordances have negative
123
Res Eng Design (2009) 20:225–239
percentage differences (again marked with red back-
grounds), such as noise and blowing dirt in front of
machine. The vacuum cleaner is in fact relatively noisy,
and does tend to blow dirt in front of the machine due to
the exhaust air flow path as directed by the motor cover.
The negative percentage differences indicate that these are
flaws with the current design that are opportunities for
redesign. This observation leads to the following method
for improving existing products using an ASM:
1. Focus on the large negative percent differences, which
is where problems are likely to occur.
2. To remedy the situation, either new parts can be
added to mitigate the negative percent difference or
Res Eng Design (2009) 20:225–239 237

Fig. 10 Five architectures for

an automotive window switch

(a) horrizontal (b) horizontal (c) lever switch (d) lever switch (e) vertical
toggle switch rocker switch on door on center toggle switch
on door on door console on door

Fig. 11 Concept selection

matrix for seven architectures
for an automotive window
switch

existing parts can be modified so they are no longer reduce the noise of the machine and susceptibility to
harmful. blocked air flow paths.
However, whenever any component is modified to
For the vacuum cleaner example, this method would
improve one affordance, any change can affect other
direct us toward redesigning the motor cover such that it
components that the original component is related to. This
would no longer blow dirt off the floor in front of the
information is captured in the top (or ‘‘roof’’) of the ASM,
machine. We would likewise be directed to redesigning
which is akin to a conventional DSM. Similarly, changing
several parts such as the motor and air paths in order to

123
238
one affordance can have an effect on other affordances.
This information is captured on the left side of the ASM.
For example, if we try to improve the ?AUA of transla-
tional movability, we are likely to simultaneously improve
the ?AAA of floor cleanability (a win-win situation, as
denoted by the strong green background). However, if we
try to improve the ?AAA of dirt removability, e.g., by
using a more powerful motor, we risk degrading the
-AAAs of loss of cleanability by blocked air flow path and
blowing dirt in front of machine (a win-lose situation, as
denoted by the orange background).
7 Using affordances in concept selection matrices
Pugh introduced decision matrices as a tool for concept
selection (Pugh 1996, p. 167–176). In the same way that we
took advantage of the property of form dependence in the
development of the ASM, we can take advantage of the
property of polarity in concept selection matrices. Recall
that the property of polarity says that affordances are either
positive or negative. In a concept selection matrix using
affordances, multiple concepts are compared against mul-
tiple positive and negative affordances. This kind of
comparison allows for each concept to be ranked in terms
of how many positive affordances each concept has, how
many negative affordances each concept has, the total
number of helpful relationships (positive affordance a
concept has plus the number of negative affordances it does
not have), and the total number of harmful relationships
(positive affordances a concept does not have plus the
number of negative affordances it does have). As with the
ASM, percentage helpful, percentage harmful, and per-
centage difference (percentage helpful minus percentage
harmful) are also calculated.
Another automotive example is shown here for illus-
tration purposes. In this instance, we are interested in
comparing seven different architectures for an automotive
window switch. The five of these architectures that have
actually been implemented are shown in Fig. 10. The
selection matrix is shown in Fig. 11.
As in the ASM, the rows of the selection matrix are
organized by ?AUA, -AUA, ?AAA, and -AAA. In the
interior of the matrix, a ‘?’ indicates that a concept has the
listed positive affordance, while a ‘-’ indicates that a
concept has the listed negative affordance. A blank entry
indicates that a concept does not have the listed affordance.
Examining the totals on the bottom of the matrix, we see
that three concepts, the horizontal toggle switch on the
door, horizontal rocker switch on the door, and vertical
toggle switch on the door all have negative percentage
difference values, indicating that they have more harmful
relationships than helpful relationships. The clear winner is
123
Res Eng Design (2009) 20:225–239
the lever switch on the center console. It has the largest
percentage difference (60%), the most positive affordances
(5 out of 5), the largest number of helpful relationships (8
out of 10), and only 2 out of 5 negative affordances and
only 2 out of 10 harmful relationships. In other work
(Gaffney et al. 2007) we have noted the historical pro-
gression of affordances over time. This trend is again
confirmed in this example, where we see the oldest archi-
tecture, the vertical toggle switch on the door, has a
negative percentage difference, and the newest architec-
ture, the lever switch on the center console, has the largest
positive percentage difference of all the architectures.
Note that with a conventional selection matrix, the
ability to distinguish between helpful and harmful rela-
tionships is lost. Note also that in this example equal
weighting is assumed for each affordance.
8 Summary remarks
In this paper several basic methods have been introduced to
allow designers to work with affordances directly. How-
ever, the methods presented herein are not intended to be
an exhaustive list of the possible ways in which affor-
dances can be dealt with practically in any design situation.
Rather, these methods are intended to serve as a starting
point for more advanced methods that may be developed in
the future. The emphasis here has been on exploiting the
unique properties of affordances in developing novel
methods.
Acknowledgments This research focuses has been supported in part
by Grant #CMMI-0826441 from the National Science Foundation.
We would like to thank Janna Sandel for her work in preparing the
vacuum cleaner example, Jonathan Thomas for his work in preparing
the automotive switch example, and Professor Gregory Mocko for his
collaboration on the automotive switch example.
References
Altshuller GS (1984) Creativity as an exact science. Gordon and
Breach Science Publishers, India. (trans. Anthony Williams)
Altshuller GS (1996) And suddenly the inventor appeared: TRIZ, the
theory of inventive problem solving. Technical Innovation
Center, Worcester, MA. (trans. Shulyak L)
Altshuller GS (1997) 40 principles: TRIZ keys to technical innova-
tion. Technical Innovation Center, Worcester, MA. (trans.
Shulyak L, Rodman S)
Altshuller GS (2000) The innovation algorithm: TRIZ, systematic
innovation and technical creativity. Technical Innovation Center,
Worcester, MA. (trans. Shulyak L, Rodman S)
Blackenfelt M (2001) Modularisation by relational matrices—a
method for the consideration of strategic and functional aspects.
In: Riitahuhta A, Pulkkinen A (eds) Design for configuration,
2001. Proceedings of the 5th WDK workshop on product
structuring, Tampere, Finland, Springer pp 134–152
Res Eng Design (2009) 20:225–239
Clausing D (1994) Quality function deployment. MIT Press,
Cambridge
Franssen M (2005) Arrow’s theorem, multi-criteria decision problems
and multi-attribute preferences in engineering design. Res Eng
Design 16:42–56
Gaffney ES, Maier JRA, Fadel GM (2007) Roles of function and
affordance in the evolution of artifacts. In: Proceedings of ICED
2007, Paris, France, Paper no. 592
Gibson JJ (1979) The theory of affordances. In: The ecological
approach to visual perception. Houghton Mifflin, Hopewell:
127–143
Hazelrigg GA (1996) Systems engineering: an approach to informa-
tion-based design. Prentice-Hall, Upper Saddle River
Hellfritz H (1978) Innovation via Galeriemethode (Innovation via the
art gallery method). Taunus, Konigstein
Kanstrup L, Lundin M (2006) Method for detection of sleepiness:
measurement of interaction between driver and vehicle. Masters
Thesis, Department of Mechanical Engineering, Linkoping
University
Kelly T, Littman J (2001) The art of innovation: lessons in creativity
from Ideo, America’s leading design firm. Doubleday, New York
Kuppuraju N, Ittimakin P, Mistree F (1985) Design through selection:
a method that works. Des Stud 6(2):96–106
Litvin FL, Fuentes A (2004) Gear geometry and applied theory, 2nd
edn. Cambridge University Press, New York
Maier JRA (2008) Rethinking design theory. Mech Eng 130(9):34–37
Maier JRA, Fadel GM (2001) Affordance: the fundamental concept in
engineering design. In: Proceedings of ASME design theory and
methodology conference, Pittsburgh, PA. Paper no. DETC2001/
DTM-21700
Maier JRA, Fadel GM (2002) Comparing function and affordance
as bases for design. In: Proceedings of ASME design theory
and methodology conference, Montreal, Canada. Paper no.
DETC2002/DTM-34029
Maier JRA, Fadel GM (2003) Affordance-based methods for design.
In: Proceedings of ASME design theory and methodology
conference, Chicago, IL. Paper no. DETC2003/DTM-48673
Maier JRA, Fadel GM (2005a) Understanding the complexity of
design. In: Braha D, Minai A, Bar-Yam Y (eds) Complex
engineering systems. Springer, New York
Maier JRA, Fadel GM (2005b) A case study contrasting German
systematic engineering design with affordance based design.
In: Proceedings of ASME design theory and methodology
conference, Long Beach, CA. Paper no. DETC2005-84954
239
Maier JRA, Fadel GM (2008) Affordance based design: a relational
theory for design, Res Eng Design (in press). doi:10.1007/
s00163-008-0060-3
Maier JRA, Ezhilan T, Fadel GM (2007) The affordance structure
matrix—a concept exploration and attention directing tool for
affordance based design. In: Proceedings of ASME design theory
and methodology conference, Las Vegas, NV. Paper no.
DETC2007-34526
Maier JRA, Fadel GM, Batisto D (2008a) An affordance based
approach to architectural theory, design, and practice. Des Stud
(submitted, under review)
Maier JRA, Sandel J, Fadel GM (2008b) Extending the affordance
structure matrix–mapping design structure and requirements to
behavior. In: Proceedings of DSM 2008, Stockholm, Sweden
Marlin R (2004) Getting geared. URL: http://www.hookmagazine.
com/septembe1.htm
Otto K, Wood K (2001) Product design: techniques in reverse
engineering and new product development. Prentice-Hall, Upper
Saddle River
Pahl G, Beitz W (1996) Engineering design: a systematic approach,
2nd edn. Springer, New York
Pugh S (1996) Creating innovative products using total design.
Prentice-Hall, New York
Rourkes N (1988) Design synectics: stimulating creativity in design.
Davis Publications, New York
Scott MJ, Antonsson EK (1999) Arrow’s theorem and engineering
design decision making. Res Eng Des 11:218–228
Shupe J, Mistree F et al (1987) Compromise: an effective approach
for the hierarchical design of structural systems. Comput Struct
26(6):1027–1037
Simon HA (1996) The sciences of the artificial, 3rd edn. MIT Press,
Cambridge
Steward DV (1981) Systems analysis and management: structure,
strategy and design. Petrocelli Books, New York
Suh NP (1990) The principles of design. Oxford University Press,
New York
Suh NP (2001) Axiomatic design: advances and applications. Oxford
University Press, New York
Webb Associates (1978) Anthropometric source book. Report num-
ber: NASA-RP-1024, S-479-VOL-1. National Aeronautics and
Space Administration, Yellow Springs
123