Managing traceability information in manufacture

Abstract

In this paper, an approach to design information systems for traceability is proposed. The paper applies gozinto graph modelling for traceability of the goods flow. A gozinto graph represents a graphical listing of raw materials, parts, intermediates and subassemblies, which a process transforms into an end product, through a sequence of operations. Next, the graphical listing has been translated into a reference data model that is the basis for designing an information system for tracking and tracing.

Materials that are modelled this way represent production and/or purchase lots or batches. The composition of a certain end product is then represented through modelling all its constituent materials along with their intermediate relations. By registering all relations between sub-ordinate and super-ordinate material lots, a method of tracking the composition of the end product is obtained. When the entire sequence of operations required for manufacturing an end product adheres to this registering of relations, a multilevel bill of lots can be compiled. That bill of lots then, provides the necessary information to determine the composition of a material item out of component items. These composition data can be used to recall any items having consumed a certain component of specific interest (e.g., deficient), but also to certify product quality or to pro-actively adjust production processes to optimise the product quality in relation to its production characteristics (e.g., scarcity, costs or time).

Introduction

Food deficiencies with important consequences have occurred over the last few decades. Food deficiencies are cause for alert, because such products have consequences with a negative impact on consumers and businesses. The consumer may be affected in his physical health and well-being, and businesses in their commercial health and well-being. In 1998, the European Commission carried out a research project that showed that 11% of all food products that are controlled by the European Union (EU) do not comply with the demands of EU legislation (Trienekens & Beulens, 2001).

It is estimated that millions of Europeans get sick every year from food contamination. Regular causes are Salmonella, Campylobacter and E. coli O 157. Moreover, it has been discovered that BSE in cattle is the probable cause for the humans deadly variant Creutzveldt-Jacob. Since 1998, there has been a large-scale crisis in the European food sector, which impacted the cattle sector. Between 1990 and 1999 there was a reduction in sales of cattle beef in the EU of 6% (on average). The British meat sector suffered the most from the crisis in this period. In 2000 several new discoveries were made in other countries such as France and Germany. By mid-February 2001, the consumption of cattle meat has dropped by as much as 80% in Germany. Clearly, products from the food industry are vulnerable.

A quote from Deasy (2002), for example, is very illustrative. “The dairy industry is under pressure to improve product security, implement efficient risk management and rapid response capabilities, and manage quality from trough to table to achieve full, verifiable traceability. In order to invest proactively in the future of the industry, individual processors must reassess their handling of quality and traceability data for strategic and competitive success, make appropriate physical configuration and operational adjustments, and deploy readily available technological tools to meet legislative and consumer demands, enhance product quality and increase operational efficiency and profits.”

Whereas Deasy stresses responsibility of the producer, in Salaün-Bidart & Salaün (2002) the roles of other agents in society is discussed, such as retailers, public authorities and consumers themselves. Additionally they state that production of data and diffusion and use of (risk) information is required to make individual choices.

Due to problems of product deficiency, one can find recall announcements to occur in newspapers; expressions of (not necessarily effective or efficient) product security and related information diffusion. In the previous subsection, it is showed that many examples of product safety issues originate from the agro- and food industry. In Abbot (1991) several cases of product recalls are described, from food industry, but also from pharmaceutical and medical products, vehicles and aerospace. A recall is a procedure to withdraw all products with a particular deficiency from the supply chain. The Food and Drug Administration (FDA) distinguishes four types of recall: recall due to violation of a law; correction, implying repair, modification, adjustment, or re-labelling of the product without its physical removal; market withdrawal due to a minor violation which does not require legal action; and stock recovery when a company corrects a product that is still under its control and has not been put on the market (Abbot, 1991).

When customers of contaminated products cannot be identified or when the product has already reached the consumer, organisations must rely on national advertisements to state their recall-message. For the enterprise, it is a challenge to correctly identify the source of a food deficiency, subsequently deduce which other products are negatively affected, and trace these products.

Whereas a recall is required in case something did go wrong, an organisation might also be keen to soothe a customer's conscience (preventively). At first glance, one may consider this a marketing issue, however a sound basis for this can be found in certified quality. Certified quality is defined as meeting the specifications that have been agreed upon in advance, and which are considered to be a particular level of quality. Certified implies that there is some level of guarantee that each of the output products meets its requirements. This is, amongst others, also recognised by Salaün and Flores (2001).

The quality specifications may refer to the product itself, but refer mainly to the production processes and production means. The requirements on processes and production means are translated into associated quality attributes of the batches of output products.

It implies that for each produced batch the relevant data are recorded, and also that all products in the batch are homogeneous in terms of the specifications. Well-known examples can be found in poultry production, where a batch is defined as follows. A flock or group of animals is characterized by a number of properties that these animals have in common. This includes the poultry house they lived in, the same genetic composition, period of the stay in the poultry house, the same conditions and care during that stay, and slaughtered together in the same slaughter process.

A further step in this problem area is pro-actively adjusting production processes. Pro-actively adjusting the characteristics of a production process might be necessary to more easily meet the specifications that are stated in quality certificates, but can also be used for the optimisation and control of processes in the organisation in general. This may include decreasing production costs, decreasing failure costs, optimisation of resources in terms of productivity and optimising product quality (van Twillert, 1999). In these cases, timely recording the properties of inputs and outputs (raw materials, semi-finished and finished products) enables dynamically controlling the process and subsequent processes in the value chain.

In the examples above it is pointed out that products can be vulnerable, which puts claims on the process with respect to recalls, certified product quality and pro-active handling in production processes. The general problem can be summarised as a requirement for traceability in manufacture. The research objective that we state here is to design a solution for information management to support traceability. The solution should be generally applicable, which means that it meets requirements from various kinds of industries. The outcome of the research, therefore, encompasses a so-called reference model. This model can be applied in real-life situations that might benefit from traceability solutions.

First, this paper describes the research approach (Section 2) and provides an overview of traceability in literature (Section 3). Next, the requirements that have been induced from the cases are explained (Section 4). Then, development of the reference data model is described by explaining the model-part of the bill of lots and/or batches (Section 5.1), the model-part of operations and variables (Section 5.2) and the integration of these two model-parts (Section 5.3). Finally, an overview of the model is given (Section 5.4).

In Section 6 we reflect on the contributions of the reference data model in relation to the problem areas this paper started with: traceability in relation to recalls, certified quality and proactive control. The paper concludes with areas for future research.