Automated Design of Modular Fixtures for Flexible

Mantffactm g Systems
M.V. Gandhi, B.S. Thompson, Industrial Technology Institute, Ann Arbor, Michigan

Abstract order for a flexible manufacturing system to be truly

flexible, all of the subsystems must be flexible, too.
The absence of adequate flexible fixturing A major impediment to achieving this high degree of
methodologies is a significant impediment to the flexibility in these modern computer integrated
evolution of truly flexible manufacturing sys-
tems. Modular fixturing is one of the principal manufacturing systems is the inability of fixturing
emerging technologies which addresses this systems to be flexible, z
void in computer integrated manufacturing. A Flexible fixturing ~-7 involves employing a
methodology is proposed herein for the auto- single fixturing system which holds parts of various
mated design and robotic assembly of these shapes and sizes when subjected to the wide variety
modular fixturing systems based on the inte-
gration of state-of-the-art methodologies from of external force fields and torques associated with
several distinctly different disciplines. The com- manufacturing operations. The advantages to be
plexity of this challenging frontier problem is gained from employing such devices are twofold.
discussed and the interdisciplinary nature of the Firstly, they would reduce the lead time and effort
solution philosophy is emphasized. associated with the design of special purpose fix-
tures; and secondly, the overhead cost associated
Keywords." Modular Fixtures, Fixturing, with storing and retrieving a multiplicity of fixtures
Modules, Flexible Manufacturing Systems,
CAD/CAM. required to effect rapid changeovers between differ-
ent manufacturing operations would be reduced.
With the market for FMSs projected to achieve
an annual expenditure of $1.8 billion by the year
Introduction 1990) and the typical cost of dedicated fixtures
amounting to 10-20% of the total system costs, 9 the
A flexible manufacturing system is defined as a technological and economic impact of flexible fix-
production unit capable of producing a range of turing technologies will clearly be of major signifi-
discrete products with a minimum of manual inter- cance. Several examples in the available litera-
vention.' It consists of production equipment work- ture t°," clearly demonstrate that adaptable fixturing
stations (machine tools or other equipment for fabri- will reduce these costs by 80%. This is truly a signifi-
cation, assembly, or treatment) linked by a materials cant saving in capital investment.
handling system to move parts from one worksta- Pioneering work in this field resulted in the
tion to another. Furthermore, it operates as an inte- development of a variety of limited special-purpose
grated system under full programmable control. In fixturing methodologies such as the petal collet, the

243
Journal of ManuJacturing Systems
Volume 5: No. 4

multileaf vise, and programmable conformable force fields must be resisted by the fixture in order
clamps? Motivated by the limited success of these that the workpiece remains in equilibrium. Since a
devices, major research institutions focused their general unconstrained spatial motion of a rigid
efforts on fundamental research-oriented investiga- body has six degrees-of-freedom, the fixture must
tions of the generic problems of flexible fixturing. impose constraints to ensure that the workpiece has
This has resulted in the evolution of a new genera- zero rigid-body degrees-of-freedom and hence,
tion of flexible fixturing methodologies which will does not move.
enhance the productivity, flexibility, and cost- The characteristics of the fixture are naturally
effectiveness of computer integrated manufacturing dependent upon the shape of the workpiece and also
systems. the manufacturing process to which it is being sub-
Modular fixtures and phase-change fixtures jected. For example, milling operations can achieve
are the two principal emerging technologies which a rapid metal removal rate with very high cutting
can fixture parts subjected to a broad spectrum of forces of a periodic nature which can excite vibra-
manufacturing tasks ranging from metalcutting tions. On the other hand, a drilling operation sub-
operations, which involve large forces and torques, jects the workpiece to a more constant level of
to smaller force field applications such as assembly, cutting forces. Hence, the design of a general
inspection, and finishing operations. Phase-change purpose fixture for machining applications is a
fixturing has been the subject of several publications complicated multiparameter optimization problem.
by the authors. 47 Furthermore, there are many differences in the
The objective of this paper is to examine the requirements for fixtures used in machining and
broad issues associated with the design, manufac- those required for assembly. For example, work-
turing, and computer engineering aspects of the pieces may be subjected to large forces and torques
former class of modular fixturing. The automated during machining operations and relatively small
design and robotic assembly of modular fixtures is forces and torques during assembly operations.
currently in an embryonic stage of development. Therefore, a fixture designed to resist cutting forces
Fully automated systems are not yet commercially will necessarily be overdesigned for use in an assem-
available for implementation in an industrial en- bly application. Consequently, two sets of flexible
vironment; however, this is the subject of vigorous fixtures will probably need to be developed; one for
research thrusts by both the academic community machining operations and the other for use in rela-
and industrial research laboratories on an interna- tively low force-field operations.
tional level. The essential ingredient for fixtures to be
employed in assembly operations is the ability to
orient and guide parts. This capability must be
Fixturing Fundamentals incorporated into the fixture without incurring the
A fixture is a device that locates and holds a penalty of a bulky structure which limits accessibil-
workpiece. '2 Although fixtures are used in most ity and dexterity.
manufacturing operations such as machining,
assembly, and inspection; attention will be con-
fined, herein, to machining applications. The
Fixture Requirements
principal considerations for this class of fixture Positive Location. A fixture must, above all
design include: positioning or locating the compo- else, hold the workpiece precisely in space to prevent
nent in the fixture; securing the component while each of the spatial movements, i.e., linear movement
machining; positioning the fixture relative to the in either direction along X, Y, and Z axes and rota-
machine tool; positioning the cutting tool relative to tional movement in either direction about each axis.
the workpiece; and minimizing the unproductive Repeatability. Identical workpiece specimens
machine time during loading and unloading. should be located by the fixture in precisely the same
When a workpiece is subjected to a machining place on repeated loading and unloading cycles.
operation, it has cutting" forces imposed upon it Rigidity. The fixture must ideally hold the
which may be mathematically defined by a resultant workpiece rigidly fixed against all external force
force and a resultant torque vector. These external fields.

244
 Journal of Manufacturing Systems
Volume 5/No. 4

Interference. The fixturing elements must en-

sure that the part is accessible to the cutting tool,
and in addition, must provide for ease of loading
and unloading. Z
Positioning F u n d a m e n t a l s . The fixture must
position the workpiece in each of three planes in
order to meet the positive location criterion. This
criterion involves defining the first plane to provide
support at three points, which is followed by assign-
ing a second plane which is perpendicular to the
first. This is accomplished by orienting the part
against any two points in the second plane. A third
plane perpendicular to each of the other two is then
defined by designating one point in it. Positive loca- o
tion of a part is guaranteed if it is in contact with
three points in the first plane, two points in the
second plane, and a single point in the third plane
(the three-two-one rule).
X
Y
In practice, the support points in each plane
are selected to be as far apart as possible in order to
ensure m a x i m u m stability. In addition, it may be
Figure 1
necessary to provide redundant supports in order to Free-body Diagram for a Fixtured Workpiece
satisfy additional constraints imposed on vibration
levels and deflections.
denoted as the y z plane. The corresponding reac-
tions in these planes are P,,, P2~, and P3, respectively.
Fixture Analysis The conditions of static equilibrium are:
3
The requirements of a generic fixturing system
r,=o (1)
to locate and hold a part have been outlined in the
i=1
previous section. Attention is now focused on
determining the force field that the fixturing system and
must provide in order to ensure static equilibrium of 3
the part being fixtured. The external forces and Y. M,:O (2)
torques imposed on the part by the manufacturing i= 1
operation, such as metalcutting for example, must
be mathematically modeled. These external disturb- where F, and M i are the cartesian c o m p o n e n t s of the
ances, in addition to the clamping and friction forces, resultant force and m o m e n t vectors.
can be expressed by a resultant load vector K' and a R : Pn + P2~ (3)
resultant m o m e n t vector M about the center of mass
of the workpiece. R:= P,, + P,z + P o
A simplified analysis for a part that has been and, m o m e n t balance,
located using the three-two-one rule serves as an
illustrative example for determining the force field 3
that must be provided by the fixture. A free-body Z (P,~a,+j.i) = M~ V i ~ {1,2,3} (4)
diagram of a part supported in a fixture is shown in j=l
Figure 1. The first plane which provides three sup- where, ali are the m o m e n t arms of P~i contributing to
port points is denoted as the xy plane, and the three m o m e n t M~.
reactions associated with the three supports are Equations (3) and (4) may be combined to yield
denoted as P , , P~2, and PI3" The second plane is the matrix equation:

245
Journal oJ ManuJacturing Systerns

[0001 °lllls0
Volume 5j No. 4

Rx classified into kits with T-slotted base plates and

0 0 0 0 0 1 PI2

l
those with a grid of holes in the base plate. The
R= 1 1 1 0 0 Pt3 former category includes kits by Halder and also the
= a41 a~z a43 a4~ a,s a,~ P21 Chinese Aeronautical Technology I m p o r t / E x p o r t
M, asi a.,2 ~/53 a54 ass a56 / /°22 Company. The latter category includes kits by
-a~, a0, a0~ a~, a05 a,~..z ',P~I BliJco Technik and Yuasa International.
Typical modular fixtures are presented in Fig-
where the components of the resultant load and ure 2 and Figure 3 from which it is evident that the
moment vectors are known quantities for each fix- flexibility of this class of devices is derived from the
turing application. The reactions Pii are the unknown large number of different configurations of the ele-
non-negative quantities which must be determined ments which may be bolted to the base plate. Each
in order to design the fixture. The solution of Eq. (5) fixture kit generally consists of a variety of element
can be accomplished using standard numerical groups such as basic carrier elements, locating adap-
methods such as Gaussian elimination. In order that ters, fasteners, and attachment elements.
a unique solution exists, it is necessary for the
determinant of the matrix to be nonsingular.
This mathematical requirement naturally
imposes constraints on the choice of support points
for the three-two-one rule. Upon determining the
support reactions, a post-processing operation must
be undertaken to calculate the deflections of the
part. The outcome of this analysis may dictate the
redesign of the fixture in order to incorporate
redundant supports.
The fixturing analysis described thus far is of a
generic nature and is therefore relevant to all classes
of fixturing systems. This analysis is especially use-
ful in the automated fixture design process since it
provides information regarding optimal prospective
locations of the fixture support points. In addition,
Figure 2
it also provides the associated reaction forces imposed Modular Fixture Incorporating a Base
on the part by the fixture. Plate with Plain and Tapped Holes
In the subsequent section, a brief description of
m o d u l a r fixtures, which are a state-of-the-art Typically, the base plates and workholding
approach to flexible fixturing, is presented prior to elements have a grid of drilled and tapped holes, as
describing the methodology for the automated shown in Figure 2, which are spaced within a toler-
design and assembly of modular fixtures for flexible ance of 0.01 mm. Hardened precision steel liners are
manufacturing systems. employed in both the base plate and the modular
components in order to ensure high accuracy. In
addition, all locating and chucking elements are
Modular Fixtures hardened and ground in order to ensure precise
This methodology, which is described in Ref- shape, accurate dimensioning, and durability. Large
erence 13, for example, is based on an extension of mounting surfaces are employed to provide high
the classic machinists' approach to developing a stability and rigidity which are crucial for precision
variety of fixtures from a combination of elements machining tasks since they can readily resist the
such as V blocks, toggle clamps, and rectangular large forces and vibrations associated with metal
blocks, which are traditionally located on a cast iron removal operations.
base plate. Clearly, the flexibility of modular fixture kits
There are several vendors of modular fixture can be exploited even in traditional modes of manu-
kits in the marketplace. These products are readily facturing practice where the fixtures are assembled

246
 Journal of Manujacturing S;stems
Volume 5 No. 4

addition, the expert system must be able to explain

the reasoning process and respond to queries per-
taining to its knowledge-base. Finally, the system
must be able to assimilate new knowledge to
enhance its decision-making capability.
The knowledge base of this expert system
would have information based on subjective infor-
mation from experts in jig and fixture design in
addition to formal fixturing rules and fundamental
theories of mathematical physics. The inference
engine, which is the brain of the expert system, must
undertake the reasoning tasks by examining the
knowledge available to arrive at reasonable subop-
timal fixturing solutions. The knowledge base and
the inference engine should be complemented by a
working memory to provide a complete structure of
the expert system. This working memory is a data-
Figure 3
M o d u l a r Fixture Incorporating a Base Plate with T Slots base for monitoring the current status of the fixtur-
ing task and other relevant information.
The process of automated design and assembly
manually. In this traditional mode, the design and of modular fixtures is initiated by abstracting the
assembly of a modular fixture can take several relevant information from the part and workpiece
hours to complete when experienced personnel in jig databases. This data is further massaged with
and fixture design are utilized. '4 The average age of information from the process planning module
such personnel in the U nited States is 55 years, and which includes the magnitudes and directions of
since this labor pool is steadily shrinking, there is a forces and torques, tool paths, and other relevant
significant incentive to analyze the procedures they constraints associated with the manufacturing pro-
adopt and store this knowledge-base in computers. cess. This domain knowledge is examined by the
Consequently, research is being undertaken to inference engine to synthesize a viable suboptimal
ensure that these essential skills are not lost. '5 This modular fixture. This typically involves the selec-
research is directed towards exploiting modern tion of surfaces to be fixtured and the associated
computer technologies, where knowledge-based module types.
expert systems synthesize reasonable configurations In addition, other pragmatic considerations
of workholding elements for various fixturing such as the need for redundant supports are also
applications. This capability is critical in automat- evaluated. A modular fixture synthesized by the
ing the modular fixturing task so that a robot can expert system in this manner satisfies the necessary
physically assemble the modular fixture for a given conditions for being a viable candidate solution for
fixturing application. fixturing. The feasibility of the candidate fixture
design can be evaluated by simulating the assembly
Automated Design and Assembly of the fixture and also the manufacturing processes
imposed on the part using solid modeling tech-
The proposed methodology for the automated niques to visually determine whether constraints are
design and assembly of modular fixtures is pre- violated.
sented schematically in Figure 4. The kernel of the This proposed computer-intensive fixture
proposed approach is an expert system which mim- design methodology involves a post-processing
ics the thought processes of specialized shop floor phase to impose a sufficiency condition on the candi-
personnel with many years of experience in jig and date fixture design solutions. Typically, this involves
fixture design. This system must have the capability employing finite element models to determine struc-
to reason with bothjudgemental and formal knowl- tural deformations and to evaluate thermal effects
edge of established phenomenological theories. In generated by the interaction of the part and the

247
Journal o t Manujacturing S) stem~
Volume5 No. 4

Workpiece fFmished Part ] ( Tool Path and Forces - I Sequence of )

Material [RequirementsJ [,in Machm,ng Operations ] LMachiningOperationsJ
Characteristics

Inference matical Models ]

Engine ~ ~ of F~xtureModules

Proposed
F~xturing
Configuration

Assembly of
Fixture on CAD
] IL Fixture
Solid Models of
Modules
System
J

-@
Suboptimal L To
Ftxture
Fixture [ v Assembly
Configuration ]Workstation

Figure 4
Flow Diagramfor the AutomatedDesignof Modular Fixtures

manufacturing process. If the sufficiency condition tion. Mathematical modeling involves the represen-
is not satisfied, the control switches back to the tation of real objects by their abstractions, which
inference engine in an effort to synthesize a fixture includes descriptions of their geometry. Task speci-
design which satisfies both the necessary and the fication is the sequence of operations that needs to
sufficiency requirements. be specified in order to realize a prescribed objective
using the geometric models developed in the pre-
Automated Fixture Design vious mathematical modeling process, The objec-
tive in fixture design is, therefore, to realize
The prospective locations of the fixture sup-
prescribed spatial relationships; namely, the relative
port points, for a given workpiece geometry, can be
positioning of the various elements of the modular
determined using the three-two-one rule, or alterna-
fixture kit. This theme is detailed in the following
tively, by utilizing the knowledge base of the expert
sections.
system dedicated to fixture design. The objective of
this design process is to synthesize a set of fixture Mathematical h4odeling. The abstract repre-
modules which, when assembled in a prescribed sentations of the mathematical modeling process
order, will satisfy the fixturing conditions. have a direct congruence to their real world equiva-
The design process comprises two distinct lents. This modeling process must inherently con-
stages: mathematical modeling and task specifica- tain the following information:

248
 Journal of ManuJacturing Systems
Volume 5/No. 4

1. Geometric descriptions of the relevant manu- plane offace, and at a normal distance equal to
facturing environment which will include the the radius of the shaft.
fixture modules, workpiece, and the cutting 3. An edge is against aface when the edge is lying
tools. in the plane of the face.
2. Physical descriptions of all objects which may 4. A sphericalface is against aface, if the center of
include material characteristics such as strength a sphere lies in a plane parallel to the plane of
and hardness. the face and the normal distance between the
3. Physical constraints which govern the admis- two is equal to the radius of curvature of the
sible motions of one module relative to another. spherical surface at the point of contact.
4. Kinematic descriptions such as rotation and 5. A vertex is against aface if it lies in the plane of
translation. the face.
The geometric description of objects is the 6. A vertex is against another vertex if they
principal c o m p o n e n t of the modeling process. This coincide.
description generally involves representing objects Similarly, a fit relationship may be defined as,
using wire frames, sculptured surfaces, and solid 1. A shaft fits a hole when their longitudinal axes
models. In fixture design, an efficient and effective lie a l o n g the same line but in opposing
representation could be realized using solid models directions.
where objects are represented as combinations of 2. Conversely, a hole fits ashaJ't when their longi-
primitive solids. The legal locations of an object are tudinal axes lie along the same line, but in the
constrained by the presence of other objects in the same direction.
environment, and the nature of the constraints A n d finally, a coplanar r e l a t i o n s h i p is
depends primarily on the shape of the objects defined as,
involved. The geometric description of the objects
1. AJace is coplanar with anotherJ'ace if they lie
is, therefore, an essential ingredient in defining
in the same plane and their normals are in the
spatial relationships for task specifications.
same direction.
Task Specifications. Task specification com-
prises the definition of a variety of model configura- The spatial relationships described so far are
tions which will realize the objectives of fixture symbolic in nature. In order to demonstrate that
design which were described earlier. At this stage of these relationships can be mathematically repre-
the design process, the location of the fixture and sented and utilized in the task specification asso-
the workpiece support are known from the fixture ciated with modular fixture design, a simple illustra-
analysis previously undertaken. In order to com- tive example is presented. '6
plete the task specification, that is, to build the Consider the task of manipulating the two
modular fixture, it is now necessary to describe the blocks shown in Figure5 such that Block 1 is against
various types of contact between two objects, which Block 2, subject to the constraint,
are governed by spatial relationships as described in
(a3 against al) A (a4 against a2)
the following section.
SpatialRelationships. Spatial relationships in
The transformation matrices associated with
terms of various geometrical features of a solid

,000]
the relevant faces of the two blocks are:
model, such as face, edge, vertex, shaft, or a hole,
can be described as against, fit, and coplanar using
symbolic terminology, t6 which is used in state-of-
the-art artificial intelligence applications. These al = 0 1 a2 = 0 1
relationships are defined as follows: 1 1 0 1
Relationship against is defined as,
1. A face is against another face when they lie in (6)
the same plane and their normals are in
opposition.
2. A face is against a shaft when the longitudinal
axis of the shaft lies in the plane parallel to the
a3 = [i°°oil [!O!]oo
0 1
1 1
a4 = 01
01

249
Journal of ManuJacturing Systems
Volume 5/No. 4

Since the rotational and the translational

components of Eq. (10) can be independently solved,
I
the rotational equation corresponding to Eq. (10)
may be simplified to yield,
twix(O,)(a2"~ttwix(-02) = M(a4'~'M. (1 1)
Block 2
Further manipulations yield,
(12)

• Y Block 1 =
[i0 0ill 0 00]
1
0
0
-1
Yl 2+z,
=
0
0
1
0
2-yz 0
0 0
-1 0
2+z2 1
The solutions to the three nontrivial algebraic
tz
equations corresponding to this matrix equation,
~ I
Block1 a r e y 2 = 1,y, - 0 and z~ = z2. The physical interpreta-
tion of this solution is that the Block 1 can only
translate along the z axis since it is constrained from
~4
translating in the x and y directions as per the task
specification. This example illustrates that the sym-
bolic relationship against can be mathematically
X represented. Clearly, other spatial relationships
a3 such as f i t and coplanar can be treated in a similar
manner. The utilization of these artificial intelli-
gence concepts in the task specification associated
Figure 5
Definition of Geometric Representations of Contacts with m o d u l a r fixture design is addressed in the next
paragraphs.
The prospective contact locations between the
In order to accomplish the specified task it is workpiece and the fixture are determined from the
necessary to undertake the manipulations of trans- fixture analysis described in an earlier section. This
lation and rotation, which are denoted symbolically information is utilized in selecting an appropriate
by trans (x, y, z) and twix (0) respectively? 7,is The fixturing module or an admissible subassembly of
two against relationships in this example may by modules, which contacts both the workpiece and the
expressed as follows: base plate in at least one point. Basic geometrical
Block I = a3 "~Mtwix(O t)trans(O,y~,zj)a I Block2 (7) feature-types such as ./'ace, edge, vertex, shaft and
hole are associated with every point of contact. The
and, objective is to determine an appropriate type of
module which will satisfy one of the spatial relation-
Block I = a4 "tMtwix(O2)trans(O,y2,z2)a2Block2 (8)
ship types such as against, f i t and coplanar.
Elimination of the c o m m o n term Block 2 from Clearly, more than one type of fixturing module
the above equation, yields: may have geometrical features that satisfy the spa-
a3" Mtwix(O t)trans(O,y, ,z~)a I = a4 "~M t wix (02) tial relationships under consideration. Module types
t r ans( O,)'z ,z2 )a2 (9) are grouped by identifying them with an informa-
tion tag which classifies them by some similarity
By defining a primed matrix to represent the such as part shape or material characteristics. Since
rotational component of a transformation, and more than one information tag is associated with
incorporating the rewrite rules of Reference 17, Eq. modules of the commercially available fixture kits,
(9) becomes: grouping techniques such as direct clustering algo-
a3"l Mtwix(O~)trans(O,y~+ l,z~+ 1) rithms, merger graphs, and group technology are
(a2")-ttwix(O~)trans(O,-y2-z2)M'~a 4 = 1 (10) employed.

250
 Journal of Manufi~cturing Systems
Volume 5/No. 4

The automated robotic assembly of modular Technology, Schlumberger Well Services, Houston,
fixtures can be undertaken by employing a task Texas. This work was funded by the Manufacturing
planner that will transform the task-level specifica- Systems Program of the National Science Founda-
tions into manipulator-level specifications. The tion under Grant No. DMC-8514871.
task-level actions are specified only by their effects
on the various fixturing elements; however, for
robotic assembly, the sequence of manipulator References
motions needed to physically assemble the modular
fixture also has to be specified. The task planner 1. M.E. Merchant. Personal Communication, October 12. 1983; cited
in Computerized Manulacturing Automation: Employment, Educa-
synthesizes the description of the geometric models, tion, and the Workplace, Washington, D.C., U.S. Congress, Office of
the manufacturing environment, the specifications Technology Assessment, OTA-CIT-235, April 1984, pp. 60-61.
2. Computerized ManuJ~tt'turing Automation: Emplo.vment, Educa-
of the robot, and other relevant information, in tion, and the Workplace, Washington, D.C., U.S. Congress, Office of
-lechnology Assessment, OTA-CIT-235, April 1984, p. 96.
order to provide an executable robot program 3. B.S. Thompson. "Flexible Fixturing--A Current Frontier in the
which physically assembles the desired modular fix- Evolution of Flexible Manufacturing Cells", ASME Paper No. 84-
WA/Prod-16, 1984.
ture configuration. 4. B.S. Thompson, M.V. Gandhi. "A Commentary on Flexible Fixtur-
ing", Applied Mechanics Reviews, Volume 39, No. 10, October 1986.
5. M.V. Gandhi, B.S. Thompson. "'Phase-Change Fixturing for
FMS", Manulacturing Engineering, Volume 93, No. 6, December 1984,
Concluding Remarks pp. 79-80.
6. M.V. Gandhi, B.S. Thompson. "'Phase-Change Fixturing for Flex-
Modular fixturing systems are one of the ible Manufacturing Systems", Journal off ManuJacturing Systems,
Volume 4, No. 1, pp. 29-39.
potentially valuable adaptable fixturing technolo- 7. M.V. Gandhi, B.S. Thompson, D.J. Maas. "'Adaptable Fixture
Design: An Analytical and Experimental Study of Fluidized-Bed Fix-
gies for flexible manufacturing systems. However, turing", A S M E Journal of Mechanisms, Transmissions, and Automa-
in order that these systems be integrated into an lion in Design, Volume 108, March 1986, pp. 15-21.
8. "'Machine Tool Makers: Healing, But Sickly", Business Week.
automated manufacturing facility, the design and October I, 1985, pp. 119-120.
assembly processes associated with their construc- 9. M. Waldish. "'On the Edge of a Snake Pit--FMSs are a Boom if
used with Careful Planning, but Few AchieveThis", Engineer, Volume
tion must be automated too. This fundamental 255, October 1982, pp. 32-33.
10. G. Lewis. "Modular Fixturing Systems", Proceedings of the 2nd
objective was addressed by proposing the integra- International Con.lerence on Fle.~:ible ManuJac'turing Systems, North
tion of state-of-the-art methodologies for the auto- Holland Publishing Company, London, United Kingdom, October
1983, pp. 451-462.
mated design and assembly tasks associated with 1 !. J.C. Quinlan. "New Ideas in Cost-Cutting, Fast-Change Fixturing",
Tooling and Production, April 1984, pp. 44-48.
synthesizing and constructing modular fixturing 12. L.E. Doyle, C.A. Keyser, J.L. Leach, G.F. Schrader. M.B. Sin~ger.
systems. Mant~/acturing Processes and Materials jbr Engineers, Prentice-
Hall, 1969.
13. H. Hammer, J. Schuster. "A New Approach for Flexible Automa-
tion in Boring and Milling Operations", Tages ZeitungJur Metallbear-
Acknowledgement beitung, Volume 77, No. 5, 1983, pp. 7-17.
14. B. Krauskopf. "Fixtures for Small Batch Production", ManuJactur-
ing Engineering, January 1984, pp. 50-53.
The authors wish to acknowledge the support 15. A. Markus. Z. Markusz, J. Farkas, J. Filemon. "'Fixture Design
and encouragement of Dr. D.R. Falkenburg, Vice Using Prolog: An Expert System", Robotics and Computer-lntegrated
Mant~lbcturing, Volume I, No. 2, 1984, pp. 167-172.
President, Research & Development, Industrial 16. M. Brady, J.M. Hollerbach, T.L. Johnson, T. Lozano-Perez, M.T.
Mason (Editors). Robot Motion: Planning and Control, The MIT
Technology Institute, and D.J. Maas, Manager, Press, Cambridge, Massachusetts, 1984.
Automated Manufacturing Technologies Labora- 17. R.J. Popplestone, A.P. Ambler, 1. Bellos. "An Interpreter for a
Language for Describing Assemblies", Artificial Intelligence, Volume
tory, Industrial Technology Institute. In addition, 14, No. I, 1980, pp. 79-107.
they wish to acknowledge the fruitful discussions 18. A.P. Ambler, R.J. Popplestone. "Inferring the Positions of Bodies
from Spatial Relationships", Artificial Intelligence, Volume 6, 1975,
with Dr. M.Y. Zarrugh, Manager, Mechanical pp. 157-174.

Author(s) Biography
Mukesh V. Gandhi is an Assistant Professor in the Department of Mechanical Engineering at Michigan
State University. He is associated with the Industrial Technology Institute in the capacity of a Senior
Research Associate. He received his Ph.D. from the University of Michigan in 1984 for work on solid-fluid

251
Journal of Manu/ac'turing Systems
Volume 5~ No. 4

interaction problems. He is the author of several papers in the areas of interacting continua, finite element
methods, high-speed mechanisms, composite materials, and flexible fixturing.
Brian S. Thompson is a Professor in the Department of Mechanical Engineering at Michigan State
University and serves as a Senior Research Associate at the Industrial Technology Institute in Ann Arbor,
Michigan. He received his BSc and MSc degrees from the University of Newcastle upon Tyne, England, in
Mechanical Engineering prior to receiving his Ph.D. in 1976 from the University of Dundee, Scotland for
research on the vibrational response of high-speed machinery. After working for the British Aerospace
Corporation for two years, he returned to academia and has since published numerous papers on mecha-
nisms, finite element methods, acoustics, machine vibrations, composite-based robotic systems, and flexible
fixturing.

252