International Journal of Machine Tools & Manufacture 43 (2003) 797809

Information model for the integration of inspection activity in a

concurrent engineering framework
J. Barreiro a,, J.E. Labarga a, A. Vizan b, J. Ros b
b

a
Department of Manufacturing Engineering, University of Leon, Escuela de Ingenieras Industrial e Informatica, 24071 Leon, Spain
Department of Mechanical Engineering, Polytechnical University of Madrid, E.T.S.I. Industriales, C/ Jose Gutierrez Abascal, 2, 28006
Madrid, Spain

Received 18 October 2002; received in revised form 12 February 2003

Abstract
The basis of the integration of activities related to the product life cycle is, mainly, the creation of a unique and coherent
information model along all the stages in the cycle. Up to date, dimensional inspection activities have not been deeply analysed,
possibly due to the fact that, their work influences are smaller than others, such us numerical analysis, materials or numerical
control. However, the integration of dimensional inspection is very important for several reasons: the need of specifying the design
and planning of the inspection process from the conceptual part design; the increase of use of high speed coordinate measuring
machine (CMM) in the production lines; and finally, for the interest of a feedback between data inspection and manufacturing
processes. This paper is a detailed proposal of an information model for inspection based on operations for CMM, which represents
a consistent structure of the necessary data in an integrated product setting.
2003 Elsevier Science Ltd. All rights reserved.
Keywords: Coordinate measuring machine; Computer Integrated Manufacturing (CIM); Information modelling; Concurrent engineering; Standard
for the exchange and sharing of product

1. Introduction
The integration of activities related to the product life
cycle is being studied and developed by several research
groups all around the world. The basis of this integration
is, mainly, the creation of a unique and coherent information model along all the stages in the cycle. The
importance of this integration and its world wide acceptance was definitively assumed in the middle of the
1980s when the STandard for the Exchange and sharing
of Product (STEP) model data initiative was launched
under the approval of ISO [1].
Due to the complexity and wide scope of the problem
to solve, different research groups are focusing on specific
stages of the production cycle but keeping the provided
partial solution under the common framework of STEP.
So far, most of the work is related to design, product data

Corresponding author. Tel.: +34-987-291792; fax: +34-987291930.

E-mail address: dfqjbg@unileon.es (J. Barreiro).

management, and Computer Numerical Control (CNC)

planning and programming. Until now, dimensional
inspection activities have not been deeply considered in
this integration effort, possibly because their influence in
the whole process is considered to be less relevant than
the design or machining activities. Refs. [24] are
examples of different work related to product integration
using STEP. A more detailed reference list of research
work can be found in the references [5,6].
However, the integration of dimensional inspection is
very important for several reasons: the need of specifying the design and planning of the inspection process
from the conceptual part design; the increase of use of
coordinate measuring machines (CMMs) in laboratories;
the increase of use of high speed CMMs in production
lines, above all in sheet forming; and finally, provides
a feedback between data inspection and manufacturing
processes. In the integrating community it is widely
recognised that the concurrent use of design using features, the application of knowledge based systems and
the modelling of information leads to a better integration
of production activities [4].

0890-6955/03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0890-6955(03)00060-9

798

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

The work presented in this paper is a detailed proposal

for an information model for inspection based on operations (inspection operation features) for CMMs. The
information contained in the data objects defined will
allow the implementation of a knowledge based system.
The decision of defining inspection operations is a
continuation of the work proposed by Ros [6] for CNC
programming based on machining operations. The proposed solution provides an inspection activity systematisation and a consistent structure of necessary data to
integrate the inspection process in a concurrent engineering environment.
2. Functional model for the inspection with CMMs
This model represents the flow of activities to be carried out during the process of inspection using a CMM.
For its development, the activities that take place along
the process have been identified and analysed, focusing
specially on the type and flow of information among
them. This functional analysis makes it possible to systematise the subsequent development of the required
data structures, which are included in the information
reference model presented further in this paper.
The IDEF0 [7] modelling technique has been used for
the functional model definition. Fig. 1 shows the first

Fig. 1.

level of activities, where four main ones have been considered: identification of elements to inspect, determination of inspection resources, detailing of inspection
plan, and execution of inspection and analysis of results.
This leaf provides a general perspective of the scope
considered in the developed work. The complete
decomposition of these main activities and their description can be found in the work by Barreiro [5].
The input and output data flow analysis in each
activity and their sub-activities enables identification of
a set of information requirements. These requirements
have been divided into 13 groups according to their
functionality, which are described in the next section.
These groups have been organised in two models: the
product model and the process model, as Fig. 2 shows.
The interaction between these two models must be
direct for the implantation of a concurrent engineering
framework. Consequently, the information structures
included in each of them must follow the same criteria
in structure and meaning. This organisation of the information groups makes easier the access and sharing of the
data related to the product and to the process, allowing a
correct application of the design for inspection concept
and the feedback of inspection results backwards the previous production stages in the product life cycle.

Information groups.

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 2.

799

Information models.

3. Information model structures: product model

and inspection model

ances (T1) and geometrical tolerances (T2). Ref. [5] contains a fully detailed version.

The information flow represented in the functional

model is the basis for the construction of an information
model, which is known as information reference model.
express Language [8], in its textual mode and the
graphical one (express-g), has been used to represent
the data structures included in this model. Several
express-g diagrams are presented in this paper; the full
version of the developed model is included in the study
by Barreiro [5].

3.1.1. Group 1DF1: product definition data

The DF1 group constitutes the core of the product
model. This group includes information about the part
definition context, composed for the generic product data
model (GPDM). GPDM is present in all the information
models related to STEP.

3.1. Description of the product model

In order to integrate this information model with other
product life cycle stages, the results achieved in the
STEP framework have been taken into account. For this
reason, data structures included in already developed
application protocols have been reviewed and basic
resources given by STEP have been added, adapting
them to the specific inspection process. The integrated
resources defined in the STEP parts 41, 42, 43 and 47
have been considered, and a full compatibility with the
AP203 and AP214 application protocols have been achieved.
The product model is composed of five information
groups: product definition (DF1), product shape (DF2),
specifications (DF3), geometry (G1), dimensional toler-

3.1.2. Group 2DF2: product shape data

This group includes information about product shape,
which can be expressed in a textual way or through symbols. In this way all the information about the part that
the designer wants to communicate can be collected and
treated in the defined product model. Table 1 shows the
basic elements considered in this information group
together with their meaning.
3.1.3. Group 3DF3: specification data
The DF3 information group contains data structures
to represent those specifications related to the part or to
the inspection process which have not been considered
in the rest of groups of the information model. Consequently, this group is a complement for the other data
groups. Although included in this section, the DF3 group
not only belongs to the product model but also to the
process model. Fig. 3 shows some basic entities defined
in this DF3 group together with their express-g representation.

800

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Table 1
Part shape information group (DF2)
Entity

Content

Product shape

Specifies the part shape using a

descriptive way, and identifies the
product definition entity related to
the shape
Identifies and defines a
characteristic portion in the part
shape, for instance one that has
associated a geometric tolerance
Identifies a relationship between
two shape aspects, for instance
the existing relationship between
an element to control and the
datum in a parallelism tolerance
definition
Associates the part shape to the
related geometry

Shape aspect

Shape aspect relationship

Shape to geometry relationship

3.1.4. Group 4G1: geometry

This group contains the necessary data structures for
the geometrical representation of the part. It has taken
into account the geometrical information entities
included in STEP which are related to the B-rep representation models.
3.1.5. Groups 5T1: dimensional tolerances and
Group 6T2: geometrical tolerances
These groups contain information about the functional
specifications established in the part with regard to tolerances. Figs 4 and 5 show some basic entities considered.
In terms of tolerances, the developed work proposes
a complementary model for those described in the Refs
[911]. Additionally, it is based on the definitions
included in the STEP standard in its 47 part, and also

Fig. 3.

on the definitions given by the ISO standards numbered

286, 1101, 2692, 5458, 5459 and 10578.
Fig. 5 shows an example of the information structure
of an entity included in this group. In particular, the perpendicularitytolerance is represented. The rest of the
entities about tolerances have a similar definition.
3.2. Inspection process model
The inspection process model has been split up into
four groups: elements to be inspected (EI1), inspection
process planning (PPI1), inspection resources (RC1) and
inspection results (RI1).
3.2.1. Group 7EI1: inspection elements
This group contains information about the part
elements to be inspected. Fig. 6 shows some data structures included in this information group. An inspection
element entity is the identification of an aspect of the part
shape which has one or more dimensional or geometrical
tolerances associated, and therefore it must be measured.
3.2.2. Group 8PPI1: inspection process planning
and Group 9DM1: measuring data
The PPI1 group contains information about the
inspection process planning, such as those related to the
process planning and to the measuring operations. DM1
group specifies the information related to contact points
on the part surfaces, which are relevant for the planning
and measuring processes and for the interpretation of
inspection results.
Entity binomials of type processplan/inspection
processplan or processoperation/inspectionprocessoperation have been used to develop the process
model. In this sense, through the definition of subtypes
of processplan and processoperation is possible to

Specifications (DF3).

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 4.

Dimensional tolerances (T1).

Fig. 5.

Geometrical tolerances (T2).

801

802

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 6.

Elements to be inspected (EI1).

integrate this model with other ones, which are related to

other activities in the production cycle and also contains
processes and operations. Fig. 7 depicts the structure of
the information Group 8.
With regard to the part alignment operation, wider
information is necessary to define its correct location in
the machine. For this reason, two subtype entities have
been defined according to the complexity of the reference surfaces.
The entity partmeasuringoperation has been
specialised into a group of basic measuring operations.
This has been done considering the geometry of the surface to measure.
With this information structure, the planning of CMM
programming can be based on inspection operations,
instead of sensor trajectories. This enables a quicker and
easier way for the definition of the global inspection process of the part. CMM control devices would accept this
information directly and generate movement orders for
the sensor. This proposal is based in the work developed
by Rios [6] for CNC, and it means the programming of
the CMM at the operation level instead of at the probe
path level defined by dimensional measuring interface
specification (DMIS) [12].
Programming based on operations instead of on trajectories has the advantage that all the information associated to each operation is visible and easily accessible,
so that it is possible to dispose of all data related to the
operation. This direct access to data makes easier to take
decisions at any point in the process. For instance, as it

is showed in Fig. 7, the information associated to a measuring operation is accessible through its attributes.
The inspection points for the measuring operation are
represented in the inspectionpointsequence entity,
which contains all the information related to the contact
point distribution, such as number of points, distribution
on the surface, approach directions, and also the criteria
followed to make these decisions. The operationcondition attribute identifies the technological parameters
associated to the machine movements, such as measuring, approach and retraction distances and speeds, etc.
Finally, the tool attribute identifies the probe assembly
used for the operation.
3.2.3. Group 10RI1: inspection results
This group contains information about the inspection
results. It includes the necessary corrective actions to be
taken on the design or on the process.
The inspection operations elementconstruction
operation and tolerancetestingoperation, are
included in this information group, Fig. 8.
The elementconstructionoperation entity applies
when a tolerance is defined over an element which is
built from other elements belonging to the nominal shape
of the part. For instance, a circle, which pass through
the centre points of several circles previously measured.
This entity has been specialised in two entities considering if the new element calculus is from simple elements
(points) or from complex elements (planes, lines, etc.).
The result of these operations is represented by means

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 7.

803

Inspection process planning (PPI1).

of an inspectionconstructedcharacteristic entity.
This entity contains the fitting algorithm and the corresponding part shape aspect.
The inspection results are represented by a tolerancetestingresults entity, which is derived from a
tolerancetestingoperation entity, where the calculus
is made.
3.2.4. Group 11RC1: inspection resources
This group contains information about the inspection
resources to be used along the process, such as CMM,
probes, and inspection fixtures.
In a first level two types of entities have been considered: cmmresource and assemblydefinition. Both
of them have been defined as a specialization of the productdefinition entity, Fig. 9. The cmmresource
entity includes all the elementary devices and tools used
in the inspection process, whereas the assemblydefinition entity is based on the representation of the existing
relationships among the components, which can configure a location fixture, a probe tool, etc.
Five types of individual resources have been identified: coordinate measuring machines, elements for the
sensor calibration, fixture elements, probe elements

(contact and no contact) and generic resources not

included in the previous cases.
The assemblydefinition entity is specialised in two
subtypes: the assemblies associated to probes and the
assemblies associated to fixtures. In the case of
assemblies, not only the contact probes have been considered but also those related to probes without any
mechanical contact (optical, capacitive, inductive, etc.).
The elements, which constitute an assembly are established through a set of entities of the component
assemblyrelationship type. The firstelement and
secondelement attributes of this entity identify the
assembly order of the components, making reference to
the relative position among them through the transformation attribute.
The express-g diagram of these entities is shown in
Fig. 9. It makes special consideration on sensor
assemblies for contact probes. It is worth mentioning the
tips attribute, which is defined as an ordered list to identify the different probe tips and their location in the
assembly (location in the list). The work direction for
every tip is defined in each of them.
The complete developed model is made up of 240
entities. The full description of all of them is contained

804

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 8.

Inspection process results (RI1).

in Ref. [5]. As an example of express textual language,

in Fig. 10 the following entities are represented: part
operation, inspectionoperation, partmeasuring
operation, inspectionoperationparameters, inspectionpointsequence, inspectionelement and real
inspectioncharacteristic.

4. Interpreted information model

The information model presented in the last section
satisfies the information requirements established from
the design activity to the inspection process, going
through the inspection planning and the feedback of the
inspection results. However, in order to integrate any
model in a global production framework based on concurrent engineering, where there are other activities and

in consequence other models are developed, it is necessary to establish a mapping between the definitions contained in the models and some kind of neutral information definitions. This correspondence between the
developed information structures and a neutral information definition is done through the use of the definitions provided by the STEP ISO standard. The STEP
definitions are contained in the standard documents
named as integrated resources. In consequence, these
resources are the base for the integration of different
information models such as the one presented in this article.
The correspondence between the model and the standard definitions is done through mapping tables. A piece
of one of the mapping tables can be seen in Table 2. In
Ref. [5] a more detailed description of these mapping
tables can be found.

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 9.

805

Inspection resources (RC1).

5. Model testing in a concurrent engineering

framework for inspection
In order to prove the usefulness of this information
model, it has been implemented in a real production
framework, where any type of data access or transference is done according to the structures included in the
product model and the process model. This framework
has been called inspection framework for concurrent
information access (IFCIA), and it is mainly composed
of three systems: product modeller, inspection equipment and product central database.
The commercial CAD/CAM software CATIA has
been used as product modeller system. This software

offers an open programming structure and an open

access to data, which has made the product modeller
development easier. A specific interface has been
developed for the generation and input of technological
and geometrical data. This interface is composed of a
functional sub-module and a user interaction sub-module. The functional sub-module defines the system architecture to manage the data structures included in the 13
groups previously defined in the information model. The
functional sub-model has been built from user application sets (UASs) and user application elements
(UAEs) concepts in the application programming interface (API) of CATIA.
On the other hand, the user interaction sub-module

806

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 10.

express information model.

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

807

Table 2
Mapping table among the information model and the STEP integrated resources
Entity

Standard resource

Inspectionpointsequence

Representation

41

Identifier
Name
Description
Pointnumber

Representation.id
Representation.name
Representation.description
Descriptiverepresentationitem

41
41
41
43

Pointnumcriterion

Descriptiverepresentationitem

43

Inspectionpointsequence to
pointwithdirectionmodel (as
pointsequence)

Compoundrepresentationitem

43

Distributioncriterion

Descriptiverepresentationitem

43

enables the data input in a semiautomatic and interactive

way, and shows the corresponding results graphically.
This module is based on programming functions offered
by the API of CATIA, particularly those included in the
IUA and CATGEO interfaces.
These two sub-modules enable the input and storage
of data according to data structures defined in the information model. However, this information is kept in the
product modeller local database. In order to let the rest
of agents in the production cycle access to these data, it
has been preferred to transfer this information to a pro-

Standard part

Mapping
Representation
{representation.contextofitems
representationcontext
representation.name=cmm inspection
point sequence representation}

Representation
{representation.name=cmm
inspection point sequence
representation}
representation.items[3]
representationitem
{representation
item.name=number of points}
descriptiverepresentationitem
Representation
{representation.name=cmm
inspection point sequence
representation}
representation.items[2]
representationitem
{representation
item.name=number of points
criterion} descriptive
representationitem
Representation
{representation.name=cmm
inspection point sequence
representation}
representation.items[1]
representationitem
{representationitem.name=point
sequence} compound
representationitem
Representation
{representation.name=cmm
inspection point sequence
representation}
representation.items[4]
representationitem
{representationitem.name=points
location criterion} descriptive
representationitem

duct central database, as Fig. 11 shows. The Express

Data Manager object oriented database has been used
for this purpose. This commercial database works using
an EXPRESS information schema. Besides these submodules, a third sub-module has been developed to
recover and transfer the information among the three
IFCIA elements.
Finally, a DEA Mistral CMM connected to a personal
computer has been used. This system receives the
inspection program based on operations from the central
database. This programme is processed in DEAPL com-

808

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

Fig. 11. IFCIA framework.

mands and it is executed in the CMM. It is also possible

to develop and map the inspection operation programme
to DMIS, in the case of machines of different manufacturers.
The inspection results are also stored in the central
database, refreshing the corresponding information entities, whereas the real geometry is also represented in the
product modeller. In this way the cycle is closed and it
is possible to establish an automatic geometrical comparison between the nominal geometry, defined in the
design activity, and the measured geometry. A more
detailed description about this framework can be found
in Ref. [5].

6. Conclusions
An information model has been developed to make
possible the integration of the dimensional inspection
process into the production cycle. The information
model is composed of a product model and a process
model, so that the relationship between both of them is
direct, and this enables the efficient link of the design
activity, the planning activity and the inspection
execution. It also makes easier the feedback of the
inspection results backwards the previous activities in
the production cycle.
The result completes current schemas and it can be
used as a basis for the development of applications
through the production cycle. This aim has been tested
through the development of an application to integrate
the CATIA system and a DEA CMM. The application

accepts and generates information according to the information structures defined in the model.
The model has been developed taking into account
inspection operations, so that it is possible to substitute
the neutral file interchange at the level of trajectories,
such as DMIS, for another defined at a higher level based
on operations. Dimensional measuring equipments
would accept this information directly, and internally,
they would make the calculus to generate the orders for
the equipment movements.

7. Acknowledgments
This work has been possible thanks to the support of
the Comisio n Interministerial de Ciencia y Tecnologa
(CICYT) through the project with reference DPI20010093-C02-01. The authors wish also to thank Paloma
Trevin o for the translation support.

References
[1] Product Data Representation and Exchange. Part 1: Overview and
Fundamental Principles, ISO 10303 Part 1, ISO, 1993.
[2] T.I.A. Ellis, R.I.M. Young, R. Bell, Modelling manufacturing
process information to support simultaneous engineering, in:
N.F.M. Roozenburg (Ed.), The Hague, Proceedings of the International Conference on Engineering Design (ICED93), 1993.
[3] T.H. Liu, An object-oriented assembly application methodology
for PDES/STEP based mechanical systems, PhD thesis, The University of Iowa, 1992.
[4] A.H. Al-Ashaab, R.I.M. Young, Information Models: An Aid to

J. Barreiro et al. / International Journal of Machine Tools & Manufacture 43 (2003) 797809

[5]

[6]

[7]

[8]

[9]

[10]

[11]
[12]

Concurrency in Injection Moulded Products Design, WAM92,

ASME, California, 1992.
J. Barreiro, Modelo Integrado de Informacio n para Inspeccio n
en Ingeniera Simulta nea, PhD thesis, Universidad Polite cnica de
Madrid, Madrid, 2001.
J. Ros, Integracio n de las funciones de programacio n de
ma quinas herramienta de control nume rico mediante una aplicacio n orientada a objetos basada en un modelo de informacio n de
operaciones de mecanizado, PhD thesis, Universidad Polite cnica
de Madrid, Madrid, 1996.
Functional Modelling Manual (IDEF0), Integrated ComputerAided Manufacturing (ICAM) Architecture, Part II, vol. IV,
1981. Air Force Wright Aeronautical Laboratories. Technical
Report AFWAL-TR-81-4023. Ohio. 45333.
ISOProduct Data Representation and ExchangePart 11: The
express Reference Language Manual, ISO 10303Part 11,
1993.
J.C. Tsai, T.C. Chuang, D.N. Guo, Development of STEP-based
dimensioning and tolerancing data model, in: Manufacturing
Science and Engineering, vol. 4, ASME, California, 1996, pp.
699708.
S. Feng, Y. Yuhwei, A dimension and tolerance data model for
concurrent design and systems integration, Journal of Manufacturing Systems 14 (6) (1995) 406426.
P. Gu, K. Chang, Product modelling using STEP, Computer
Aided Design 27 (3) (1995) 163179.
ANSI/CAM-I Dimensional Measuring Interface Specification,
Computer Aided Manufacturing International, ANSICAM-I,
Arlington. Texas, USA, 1990 (101-1990).

809

Dr JOAQUIN BARREIRO is professor of manufacturing engineering at

the University of Leo n (Spain). Prior to joining, he enjoyed 2 years a grant
for researching at the University of Cantabria (Spain). His area of expertise
is in manufacturing systems and processes, with particular emphasis on
computer integrated manufacturing and concurrent engineering. He
received his PhD degree from Politechnical University of Madrid in 2001.
He is a Fellow Member of ASME International, Spanish Society of Mechanical Engineering and Spanish Society of Robotics and Automation.
Dr JULIO LABARGA is professor of manufacturing engineering at the
University of Leo n and manager of the Automatic and Manufacturing
Research Institute of Leo n. Prior to joining University of Leo n in 1989,
he was safety chief at Antibioticos S.A. (Montedison Group). His area of
expertise is in manufacturing processes, with particular emphasis on metal
cutting processes and inspection, and quality. He is a member of ASME
International and the Spanish Society for Quality.
N is professor and chief of the manufacturing enginDr ANTONIO VIZA
eering group at the Politechnical University of Madrid. Prior to join this
university he was working as professor at the University of Oviedo
(Spain). His publications spans the areas of metal cutting, numerical control, robotics, and others. Now he is working in the polymer processing
area.
Dr JOSE RIOS is professor of manufacturing engineering at the Politechnical University of Madrid (Spain). He received his PhD degree from
Politechnical University of Madrid in 1996. His area of expertise is in
manufacturing systems and processes, with particular emphasis on information modelling and concurrent engineering. He has collaborated with
workgroups for the development of STEP standard.