A Food Value Chain Integrated Business Process and Domain Models for
Product Traceability and Quality Monitoring: Pattern Models for Food
Traceability Platforms
Estrela Ferreira Cruz
1,2
and Ant
´
onio Miguel Rosado da Cruz
1,2
1
ARC4DigiT - Applied Research Centre for Digital Transformation, Instituto Polit
´
ecnico de Viana do Castelo, Portugal
2
Centro ALGORITMI, Escola de Engenharia, Universidade do Minho, Guimar
˜
aes, Portugal
Keywords:
Business Process Modeling, BPMN Process Model, Domain Model, Value Chain Integration, Perishable
Products, Food Products, Quality Monitoring and Tracing, Product Lot Localization Traceability.
Abstract:
Traceability of product lots in perishable products’ value chains, such as food products, is driven by increasing
quality demands and customers’ awareness. Products’ traceability is related to the geographical origin and
location of products and their transport and storage conditions. These properties must be continuosly measured
and monitored, enabling products’ lots traceability concerning location and quality throughout the value chain.
This paper proposes pattern integrated business-process and domain models for food product lots traceability
in the inter-organizational space inside a food value chain, allowing organizations to exchange information
about the quality and location of product lots, from their production and first sale until the sale to the final
customer, passing through the transportation, storage, transformation and sale of each lot. The paper also
presents the process followed for obtaining these two pattern models. Three exploratory case studies are used,
towards the end of the paper, for validating the proposed business-process and domain pattern models.
1 INTRODUCTION
Consumers have a growing interest in knowing ev-
erything about the products they consume, and have
more trust in brands that have a tight control of their
products’ quality. They also want to know the origin
of the products they are buying or eating, and where,
how and in what conditions products are transported
and stored. In what concerns fresh products, however,
it may not be easy to know the quality or origin of
what they are buying. Collecting quality control and
traceability data from all steps of a fresh foods value
chain, from product’s harvesting (capture, creation) is
a complex task involving many business partners.
Organizations in a particular value chain exchange
products, documents, and information that allow them
to add value to the products they buy before trans-
forming and/or selling them to the next link in the
value chain. Internationalization and digital transfor-
mation demand from a value chain’s operators an in-
creasing interconnectivity and integration. Addition-
ally, in perishable products’ value chains, such as food
(e.g. fishery, meat, milk and dairy, fruits) or phar-
maceutical products, increasing quality demands and
customers’ awareness entails knowing the geograph-
ical origin of products and that products location and
their transport and storage conditions are continuosly
measured and monitored. While the geographical ori-
gin is most important for fresh or transformed agri-
cultural, dairy products, etc., especially when talking
about products with protected designation of origin,
transport and storage conditions are important both
for perishable food products, and for other products
that bear a validity date, including pharmaceutical
products. Products’ lots localization is also very im-
portant, especially when public health may be at stake
(remember the cases of mad cow disease or african
swine fever) and products’ lots must be recalled.
The increasing interconnectivity of a value chain’s
operating organizations demands better integration of
their processes, either by allowing business partners
to interact with the systems that support the processes
through an external business partners user interface
(Cruz and da Cruz, 2018), or by directly integrating
the companies processes and systems that create prod-
uct information and enable the exchange of informa-
tion. The supply-chain oriented business-to-business
(B2B) systems’ integration, aiming the interchange of
information and documents about trading, payments,
etc. is thoroughly covered by the state of the art.
Cruz, E. and Miguel Rosado Cruz, A.
A Food Value Chain Integrated Business Process and Domain Models for Product Traceability and Quality Monitoring: Pattern Models for Food Traceability Platforms.
DOI: 10.5220/0007730502850294
In Proceedings of the 21st International Conference on Enterpr ise Information Systems (ICEIS 2019), pages 285-294
ISBN: 978-989-758-372-8
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
285
These B2B oriented integration solutions are not suit-
able for traceability and quality monitoring purposes,
though, because of the following reasons:
Information about products localisation is not typ-
ically maintained in any of the value chain’s oper-
ators systems. This makes impossible for compa-
nies to be sure about the origin of a product, and
impedes the knowledge of the locations to where
a lot has been splited and distributed, when lot re-
trieval from market (lot recall) is needed.
Products quality information is registered in each
value chain operator (typically handwritten on a
notebook), but is never shared among business
partners. This makes impossible for a company to
be sure that the products it received under appar-
ently acceptable environmental conditions, have
not been subjected to product spoiling or deterio-
ration conditions at any point in the value chain.
Food traceability is of utmost importance, namelly
as it allows to avoid forgeries, by assuring the origin
of a product, especially when talking about products
with protected designation of origin; and, it enables
to recall products’ lots, because of food contamina-
tion or other threats to the public health. Food trace-
ability requires a common value chain’s platform that
enables sharing information about products and their
agreed quality parameters along the value chain.
In these cases, traceability information will en-
able identifying the origin of each product lot and all
the geographical location points where it has passed,
along with information about the packaging, transport
and storing conditions (e.g. temperature, humidity),
and about the observed quality of the product at each
point. This grants greater security to consumers.
The main contribution of this paper is the presen-
tation of a pattern business process model for perish-
able products value chains, and the corresponding do-
main entities model. These models fill the gap space
that lies between operators in a value chain.
The structure of presentation is as follows: In the
next section, related work is presented. In section 3,
a value chain’s integrated business process model is
presented as a pattern for perishable products value
chains. Section 4 presents the corresponding pattern
domain model. Section 5 arguments towards valida-
tion of the results obtained and section 6 concludes
the paper and discusses directions for future work.
2 RELATED WORK
In 2002, the European Union created a directive re-
garding the traceability in food sector, assuring in-
formation flow transparency and traceability (Regu-
lation (EC) No 178/2002). This directive defines a set
of principles, requirements and procedures in matters
of food and feed safety, covering all stages of food
and feed production and distribution (UNION, 2002).
Since then several authors proposed frameworks to
support the traceability in feed and food value chains.
ISO 22000 series on food management systems
also addresses the safety of food products. ISO
22005:2007 Traceability in the feed and food chain
General principles and basic requirements for system
design and implementation, is the most recent series
of food safety standards. ISO 22005:2007 gives the
principles and specifies the basic requirements for the
design and implementation of a feed and food trace-
ability system (22005:2007, 2007).
In (Regattieri et al., 2007) the authors propose
a platform to support the traceability of the famous
Italian cheese Parmigiano Reggiano from the bovine
farm to the final consumer. The framework supports
the identification of the characteristics of the product
in its different aspects along the value chain: bovine
farm, dairy, seasoning warehouse and packaging fac-
tory. The system developed is based on a central
database that collects data in all identified steps in the
food chain. Some of the information is automatically
collected by using sensors, bar codes, etc. Other in-
formation is collected manually.
In (Ioannis Manikas, 2009) the authors present the
first step in a project to design and create a Web plat-
form to support food traceability for dairy products.
The first phase of the project includes the design of a
reference model for a generic dairy supply chain. The
authors identified three main phases in a supply chain,
namely natural environment, transformation and dis-
tribution. The authors also identify the main entities
involved (actor, container, material and sample) and
present a class diagram for each entity.
Folinas et al. propose a web application for mod-
eling agricultural processes and data. The proposed
application highlights the collaborative effort in mod-
eling and integration of logistics processes assuring
that all business partners have access and share infor-
mation (Folinas et al., 2003). For that, they define
standards and specify the exchanged information.
Dabbene and Gay present new methods for mea-
suring and optimizing the performance and cost of
traceability systems by using graphs. A food produc-
tion process is seen as a sequence of storage/carrying
actions and of unit operations. A unit operation repre-
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
286
sents a product at a time. Each container/processing-
unit that individually stores/processes a product, at
a certain time, is modeled as a node in a graph
(Dabbene and Gay, 2011).
Palma-Mendoza and Neailey proposed an ap-
proach to integrate the business processes of the en-
tities involved in a supply chain. The main aim is
to guide the business process redesign to support e-
business, thus focusing supply chain B2B integration
and not a value chain oriented organizational collabo-
ration (Palma-Mendoza and Neailey, 2015).
Bevilacqua et al. propose a business process
reengineering for a supply chain of vegetable prod-
ucts (fresh vegetables, canned vegetables, mixed veg-
etables, cooked and pre-cooked vegetables) (Bevilac-
qua et al., 2009). They also suggest software system
design models for managing product traceability. The
authors present a framework based on EPCs (event-
driven process chains) and use ARIS tool to create a
Web interface to provide information to the final con-
sumer (Bevilacqua et al., 2009).
Meroni et al. propose an approach to integrate
and coordinate multi-party business processes and
present a prototype to demonstrate the proposed ap-
proach (Meroni et al., 2018). The paper presents a
methodology to translate BPMN processes to E-GSM
(Extended-Guard-Stage-Milestone) notation.
All the mentioned approaches propose specific so-
lutions for specific problems, and only a few present
specific business process or domain models. Here, a
pattern solution is proposed for any food value chain
traceability problem, namely a pattern business pro-
cess model and the associated pattern domain model.
3 MODELING THE INTEGRATED
BUSINESS PROCESS
This section presents the creation of the integrated
business process model for traceability and quality
monitoring of food products.
After having visited several producers, small
farmers’ associations, food processing industries, dis-
tributors, and supermarkets, and having participated
in the identification and design of several food chain
operators’ product traceability processes, of which
two are presented here, the similarity of different food
value chains, in terms of activities needed to address
traceability and quality monitoring, became apparent.
The approach consisted in eliciting and analyzing
a set of productive business process models from dif-
ferent food value chains’ operators (milk and dairy,
fruit, fisheries, vegetables), understand their occur-
ring order in the chain and integrate them, to obtain
a pattern business process model for traceability and
quality monitoring in food value chains.
To model business processes we are using BPMN
(Business Process Model and Notation), which is the
main standard process modeling language and one of
the most used for that purpose (Cruz et al., 2014).
Created by OMG for providing a notation clear to all
stakeholders involved in Business Process Manage-
ment (BPM), BPMN is easy to understand and usable
by people with different roles and training from top
managers to IT professionals (OMG, 2011), and is ac-
tually used both in academia and in organizations.
Business Process (BP) model diagrams define a
set of business activities carried out by an organiza-
tion for the attainment of a goal (product or service).
This type of model describes a BP internal to a spe-
cific organization(OMG, 2011). Yet, in this paper, we
do not use BPMN in the “usual” form. In fact, the
BPMN business process model is being used as a sim-
ple and easy way to understand and identify all the
activities that affect a product lot (or batch). So, in
this paper, the main lane (in the main pool) does not
represent nor a company neither a business partner.
Instead, it represents a product lot. So, business pro-
cesses are used here to focus our attention in the ac-
tivities that involve, or affect, a products lot and about
which a traceability platform needs to receive and
store information. External participants, in the mod-
els, represent the value chain operators responsible for
providing information about the activity with which
they are exchanging messages. These messages rep-
resent the information that a participant needs to pro-
vide. The activities are, in fact, executed by the par-
ticipants, represented as external participants sending
messages to the corresponding value chain activity.
The traceability platform’s main goal is to gather
and store all the relevant information from value chain
activities, to be able to identify: what was done; who
did it; when it was done; where it was done; under
what conditions it was executed. Thus, all operators
involved in the value chain must be identified.
For this presentation, and due to space limitations,
we have selected two different flows of activities in
food value chains. The next subsections present two
value chain business process models, respectivelly
about fresh vegetables, in subsection 3.1, and freeze-
dried apples, in subsection 3.2. Finally, subsection 3.3
presents a process model integrating all food value
chain’s activities, not just from the two cases pre-
sented, but also from case studies from other food
value chains (e.g. fishery, meat, aquaculture, dairy).
A Food Value Chain Integrated Business Process and Domain Models for Product Traceability and Quality Monitoring: Pattern Models for
Food Traceability Platforms
287
Figure 1: Value chain process for fresh vegetables.
3.1 Fresh Vegetables
Fresh vegetables, like fresh sardines, fresh milk or
other fresh food products, have several conditioning
factors for transport and storage. The fresh vegeta-
bles BP model is represented in Figure 1. In a value
chain of fresh vegetables, major producers, typically
owning their own brand, register and control the in-
formation about harvesting (activity Production in the
Figure 1), and assess, control and register the products
quality (Registration and Quality Assessment) before
selling them (Sell) to retailers. In these cases, the pro-
ducers are responsible for providing all that informa-
tion. Smaller producers, typically deliver their prod-
ucts in an agricultural cooperative or producers’ as-
sociation, which is responsible for registering infor-
mation about producers, products, and quality assess-
ment, and to sell products to retailers.
After being sold, products’ lots are usually trans-
ported to other sites. The information about who
transports and under what conditions the product is
transported must be stored. This information may be
provided by the transporter itself, but typically it is the
operator who receives the product that evaluates the
products’ lots quality conditions after the transport.
When products’ lots are received by a new owner,
this registers and assesses the products’ quality, be-
fore storing, transforming, or selling them. If, any-
time, a product’s lot quality is not acceptable, that
lot is downed. Information about storage conditions
(dates, cleaning conditions, temperature, etc.) must
also be registered. Products sold to the final customer
end their path through the value chain.
3.2 Freeze-dried Apples
Freeze-drying is a form of drying that removes all
moisture with almost no effect on a food’s taste. In
freeze-drying, food is frozen and placed in a strong
vacuum. The water in the food then sublimates. This
example deals with freeze-dried apples. The process
begins in the orchard in the production of apples (see
Figure 2). Once again, there are big and small produc-
ers. Major producers negotiate directly with industry,
while smaller ones deliver their products in associa-
tions or farmers cooperatives, responsible for control-
ling the quality and negotiating with industry. After
being sold, the apples are transported. The conditions
under which they are transported are always checked.
In any case, whenever apples arrive in the freeze-dried
factory plants, apple lots are received and registered,
and the quality is checked again before being submit-
ted to the transformation. A set of activities are exe-
cuted to obtain the final product. First the apples are
washed, then are sliced and freeze-dried, then pass
through vacuum and finally are packed. After packag-
ing, packet lots of sliced freeze-dried apples are stored
and wait their turn to be sold.
For enabling traceability, it must be possible to
know which new lots of freeze-dried apples came
from which previous lots of fresh apples. In other
words, lots of fresh apples may be transformed into
one or more lots of freeze-dried apples. In fact, a
given lot may be partially sold for a retailer, and par-
tially sold for a restaurant, and even partially sold
for being transformed. And industrial transformation
processes, such as freeze-drying apples, will create
lots of products from previous lots of fresh apples.
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
288
Figure 2: Value chain process for freeze-dried apples.
3.3 Food Value Chain’s Integrated
Process
The model of a BP at value chain level, represented in
Figure 3, serves as a pattern for representing any food
value chain inter-organizational collaboration space,
for enabling the traceability and quality monitoring of
products. In this BP model, six main activities have
been identified: Production, Registration and Quality
Assessment, Sell, Store, Transportation and Transfor-
mation. Additionally, there is activity Down, which is
executed if the product lot is spoiled or deteriorated,
or its validity date is expired.
With the exception of the Production and Down
activites, which may only happen once per each prod-
uct lot, every other activity may happen several times
in the product lot lifespan.
The value chain starts in the producer, where in-
formation about production (fishing, harvesting, fruit
picking, etc.) must be gathered. This information may
be provided by the producer itself (major producers)
or by associations of producers or agricultural coop-
eratives for small producers. Products are, then, reg-
istered and have their quality assessed.
As we may see in the process represented in Fig-
ure 3, after the registration and quality assessment, a
product lot may be sold or stored or transformed. Af-
ter each of these activities, the product lot is received
and its quality assessed. It may stay within this itera-
tion of activities during some time.
When a product lot is subjected to various pro-
cessing stages, the quality is verified at the end of
each stage (transformation task). The quality of the
product is assessed every time the product suffers a
transformation, a storage or a transportation. Some
of those times new lots are created and registered (as
is the case of transformation). Usually, after being
stored, a lot of product is sold. And, after being sold,
if the buyer is the final consumer, the process ends,
otherwise the product may be transported to the pur-
chasing organization, and that event must be regis-
tered for tracing purposes.
Transformation processes, together with the iden-
tified fresh products value chain’s activities, are rep-
resented in the Food Value Chain Integrated Business
Process for product traceability and quality monitor-
ing (see Figure 3). This comprises the value chain ac-
tivities (corresponding to value chain operators’ busi-
ness processes) of both the fresh products and the
transformed products. Note that, despite each value
chain activity corresponds to a business process in
the value chain operator that is responsible for it, in
the interorganizational traceability business process
model, in Figure 3, each activity refers to the gath-
ering of information from the process with the same
name. It is this information, gathered in a common
shared integration platform, that will enable prod-
ucts traceability and quality monitoring along the full
value chain path of each product lot.
Despite the traceability and quality data being di-
rectly persisted in the interorganizational space in the
A Food Value Chain Integrated Business Process and Domain Models for Product Traceability and Quality Monitoring: Pattern Models for
Food Traceability Platforms
289
Figure 3: Value chain integrated process model.
Figure 4: Derived domain model.
value chain, the data itself must be communicated by
the involved chain operator’s relevant process. Thus,
each activity can be seen as a value chain event.
4 DOMAIN ENTITIES MODEL
This section presents the domain model to support the
integrated process for traceability and quality moni-
toring in food value chains. In the next subsection,
we apply the approach presented in (Cruz et al., 2012;
Cruz et al., 2015) to derive a default domain model
from the integrated business process model presented
in subsection 3.3. Then, in subsection 4.2 we refine
the domain model to obtain a pattern domain model
that supports integrated platforms for traceability and
quality monitoring in food value chains.
4.1 Deriving the “Default” Domain
Model
In (Cruz et al., 2015) an approach has been presented,
to derive a domain (data) model by aggregating all the
information about persistent data that can be extracted
from business process models. In summary, in that
approach (Cruz et al., 2012; Cruz et al., 2015):
A participant in the business process model gives
origin to an entity in the domain model;
A data store also gives origin to an entity;
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
290
Figure 5: Refined domain entities model.
The relation between entities is derived from the
information exchanged between participants and
the activities that operate the data stores, and from
the information that flows through the process.
The entity that represents the participant repre-
sented by the main lane (pool) is related with
all entities in which information is stored (repre-
sented as data stores).
Entities representing external participants that
send messages to activities that store information
in a data store are related with the entity that rep-
resents that data store.
By applying the approach presented in (Cruz et al.,
2012; Cruz et al., 2015) to the final integration busi-
ness process model (represented in Figure 3) we ob-
tain the domain model presented in Figure 4. In this
derived domain model, the entity Lot/Batch is de-
rived from the main lane (Products lot), after renam-
ing. The entities Producer, Industry/Retailer, Logis-
tics Company are derived from the external partici-
pants with the same name. The entities Production,
Quality Assessment, Sale, Lot, Storage, Transport
and Transformation are derived from the data stores
with the same name.
Messages sent by external participants to the ac-
tivities represented in the main lane, represent the in-
formation received by the activity that needs to be
stored (represented by the data store). This way, in
the domain model, the entity that represents the data
store is related with the entity that represents the ex-
ternal participant that sends the message (Cruz et al.,
2012; Cruz et al., 2015). Following this approach:
A Producer, responsible for a Production, sends a
message to the activity that stores Production in-
formation, so entities Producer and Production are
related. The Producer may execute this process
several times so the relation is one to many.
Store, Sell and Transformation are responsibility
of Industry/Retailers;
Transport information is sent by the logistics com-
pany (or by the truck or driver), so entities Trans-
port and Logistics Company are related.
Both logistics companies and Industry (tranfor-
mation companies, such as canning industries)
send information about quality assessment, stor-
age and sales, so entities Logistics Company and
Industry are related with all those entities.
Entities Production, Transport, Storage, Sale,
Transformation and Quality Assessment are re-
lated with entity Lot/Batch because the activities
that store information in the corresponding data
stores are executed inside the main lane (in the
main pool) and according to (Cruz et al., 2012;
A Food Value Chain Integrated Business Process and Domain Models for Product Traceability and Quality Monitoring: Pattern Models for
Food Traceability Platforms
291
Figure 6: Fresh Apples from Small Farmer to Big Retailer.
Cruz et al., 2015), when a participant is respon-
sible for an activity that writes information in a
data store, the entity that represents the participant
must be related with the entity that represents the
data store.
The information about transport, quality assess-
ment, sale and transformation may be stored any num-
ber of times and the information may be provided by
different value chain operators.
The next section refines this model for a better
object-oriented structuring, and with reusability in
mind.
4.2 Refining the Domain Model
The derived domain model (Figure 4) is a default
model that may be now used for refinement into
a complete and fully understandable object-oriented
domain model. This section illustrates the refinement
of the derived model.
First, let’s remember that:
1. Producer, Industry, Retailer and Logistics Com-
pany are all Value Chain Operators;
2. Production, Quality Assessment, Sale, Storage,
Transport and Transformation are Value Chain
Events (that gather traceability and quality infor-
mation about product Lots);
3. Small producers do not register their production.
Only when delivered to a farmers association or
cooperative the production is wheighed and the
product lot is registered.
4. When receiving fresh products, some value chain
operators wheight and assess the quality of re-
ceived products and create/register a new product
batch, with a new reference or Lot number. This
new lot reference must be associated to the previ-
ous lot reference, for traceability.
5. After a transformation process, new lots of prod-
ucts may be created, referring to the new trans-
formed product.
6. Lots are allways about a product. E.g.: Lot of
fresh sardines, canned sardines, frozen sardines.
From item 1, above, in the refined Domain Model
(Figure 5) we can then find a ValueChainOperator
entity, which is an abstract class representing any
value chain operator: Producer, Industry, Retailer
and Logistics Company. These are then subclasses
of ValueChainOperator (not represented in the di-
agram). We have, also, from item 2, an Event en-
tity, which is an abstract class representing any event
about Lots in the value chain. By value chain event,
we mean any activity that leaves a trail of information
on the platform. Namely, Production, Quality Assess-
ment, Sale, Storage, Transport and Transformation,
which are represented as subclasses of Event.
In the figure, originLots represents the lots that
are associated to the event. A ProductRegistration
is where a value chain operator (a Producer, which,
by 3 may be a major producer or a producers’ as-
sociation) typically creates a new Product Lot. It
doesn’t have an origin lot, having a producedLot in-
stead. Additionally, there are events that may im-
ply the creation of new lots from previous ones,
namelly Trans f ormation (from item 5, above) and
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
292
QualityAssessment (from item 4). These events
may be associated to the set of newly created lots
(destLots). In the refined diagram (Figure 5),
the many-to-many relation between Lot/Batch and
Trans f ormation, from the diagram in Figure 4,
has given place to a many-to-many relation from
Lot/Batch to Event (a lot may have many events;
and, an event may have many original lots), and a
one-to-many from Trans f ormation to Lot/Batch (a
transformation may create several new product lots).
Event QualityAssessment has been given the same lot
association properties as Trans f ormation. All other
events have one and only one original Lot.
A product lot is always about a product (see item
6 above, and entity Product in the refined diagram).
A Lot’s Sale or Transportation may be associ-
ated to a destination value chain operator (destOper).
A Lot may yet be stored (event Stor age) during an
amout of time or it may be shut down (Downed) if it
is deteriorated or its validity date has been reached.
The diagram in Figure 5 presents only the main
attributes of each entity class. The real appropriate
attributes depend from case to case, from which prod-
ucts’ lots properties are being monitored and traced in
each case. The next section presents three case stud-
ies for validating the presented business process and
domain models.
5 VALIDATION
In this section we validate our resulting models
through three exploratory case studies. Figures in this
section are organized into three lines where, the first
line contains the events that happen, from the value
chain integrated process. The second line explains the
case study situation from the point of view of each
operator. And, the third line shows an object dia-
gram with objects created by the traceability platform,
when the events happen.
Fresh Apples from Small Farmer to Small Re-
tailer. In the first case, a small farmer produces
apples and delivers them to a farmers’ cooperative.
The cooperative, then, registers the production lot, as-
sesses the products quality and packs the apples for
being sold and transported to retailers. After being
received at small retailers’ stores, they put the apples
ready for sale. For reasons of space and simplicity of
the case, there is no figure for this situation. Here, the
original Lot makes its way until the final consumer.
At any point, it is possible to know the information
about the lot and all of its events, as long as the lot’s
packages at the several retailers bear the lot reference
or number, because all the events are linked to the
original lot.
Fresh Apples from Small Farmer to Big Re-
tailer. Figure 6 shows a similar case, but having
the apples’ lot sold to big retailers. These, typically
create lots in their ERP system, when ordering prod-
ucts from the suppliers. And then, when receiving
the supplied lots, they re-assign them to the lots they
had created before. In this case, then, when being re-
ceived by the retailer, the traceability platform instan-
tiates class QualityAssessment which allows to create
a new lot, enabling to trace between the original lot
and the newly created lot. After this event, all subse-
quent value chain events must refer to this new lot.
In this case, it is possible to know the information
about the last lot and all of its events because all the
events are linked to that lot. As events have a times-
tamp (dateTime), it is also possible to order the events
of the last lot and see that it has been created in a
QualityAssessment event associated to another origi-
nal lot, and, from there, trace all the information about
the original lot.
Fresh Apples from Small Farmer to Freeze-
drying Industry. In Figure 7, the situation depicted
refers to when fresh products are sold to industries
for being transformed. After the first sale, prod-
ucts are transported to a factory and their quality
is assessed and registered in the object instance of
QualityAssessment. Then, after being submitted to
the industrial process, the resulting lot is registered
and has its quality assessed. This creates an instance
of Trans f ormation, which alows to associate the re-
sulting lot to the previous lot of the original fresh
product. In this case, the resulting lot will refer to
a new product (freeze-dried apples, not fresh apples).
In this case it is possible to trace back to the original
lot, from the reference of the last created lot.
6 CONCLUSIONS
For coping with the growing need of food prod-
ucts traceability and quality monitoring, this paper
presents a pattern business process model and a corre-
sponding pattern domain model derived from a set of
situations observed in food value chains. In the pro-
posed food value chain business process model, seven
activities have been identified. These activities cor-
respond to events in the value chain where informa-
tion must be delivered for an integrated traceability
platform, which lies in the space between value chain
operators. This information enables to know when,
where and how each product lot has been treated and
who participated in the process.
The proposed domain model has been obtained
A Food Value Chain Integrated Business Process and Domain Models for Product Traceability and Quality Monitoring: Pattern Models for
Food Traceability Platforms
293
Figure 7: Fresh Apples from Small Farmer to Freeze-drying Industry.
from the pattern BP model in two steps. First, a de-
fault domain model has been derived from the pattern
BP model by using the approach presented in (Cruz
et al., 2012; Cruz et al., 2015). Then, the derived
domain model has been refined for improving model
structure and reusability.
Three case studies have been presented to illus-
trate the approach. The two pattern models may now
serve as template models for designing traceability
platforms for food value chains. The entities’ at-
tributes, in the domain model (Figure 5), are not com-
plete. Their completion depends on the information
that needs to be monitored and traced, in each real
case traceability scenario.
The models obtained in this work are being used
and adapted to a fishery products value chain, for
quality monitoring and traceability. Future work will
also apply these to other products value chains.
REFERENCES
22005:2007, I. (2007). Traceability in the feed and food
chain. Technical report, Int’l Standards Organization.
Bevilacqua, M., Ciarapica, F., and Giacchetta, G. (2009).
Business process reengineering of a supply chain and
a traceability system: A case study. Journal of Food
Engineering, 93(1):13 – 22.
Cruz, E. F. and da Cruz, A. M. R. (2018). Deriving inte-
grated software design models from BPMN business
process models. In Proceedings of ICSOFT 2018 - Vol
1, pp 571–582. INSTICC.
Cruz, E. F., Machado, R. J., and Santos, M. Y. (2012). From
business process modeling to data model: A system-
atic approach. In QUATIC 2012, Thematic Track on
Quality in ICT Requirements Engineering, U.S.A., pp
205–210. IEEE Computer Society.
Cruz, E. F., Machado, R. J., and Santos, M. Y. (2014).
Derivation of data-driven software models from busi-
ness process representations. In 9th International
Conference on the Quality of Information and Com-
munications Technology (QUATIC2014), pages 276–
281. IEEE Computer Society.
Cruz, E. F., Santos, M. Y., and Machado, R. J. (2015). De-
riving a data model from a set of interrelated business
process models. In 17th International Conference on
Enterprise Information Systems, pages 49–59.
Dabbene, F. and Gay, P. (2011). Food traceability systems:
Performance evaluation and optimization. Computers
and Electronics in Agriculture, 75(1):139 – 146.
Folinas, D., Vlachopoulou, M., Manthou, V., and Manos, B.
(2003). A web-based integration of data and processes
in the agribusiness supply chain. In EFITA conference.
Ioannis Manikas, B. M. (2009). Design of an integrated sup-
ply chain model for supporting traceability of dairy
products. International journal of dairy technology.
Meroni, G., Baresi, L., Montali, M., and Plebani, P. (2018).
Multi-party business process compliance monitoring
through IoT-enabled artifacts. Information Systems,
73:61 - 78.
OMG (2011). Business process model and notation
(BPMN), version 2.0. Technical report, OMG.
Palma-Mendoza, J. A. and Neailey, K. (2015). A busi-
ness process re-design methodology to support sup-
ply chain integration: Application in an airline MRO
supply chain. International Journal Information Man-
agement, 35(5):620-631.
Regattieri, A., Gamberi, M., and Manzini, R. (2007). Trace-
ability of food products: General framework and ex-
perimental evidence. Journal of Food Engineering,
81(2):347 – 356.
UNION, E. (2002). Regulation (ec) no 178/2002 of the eu-
ropean parliament and of the council. Technical re-
port, Official Journal of the European Communities.
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
294