Food Engineering Innovations Across the Food Supply Chain

By Pablo Juliano and Minh H. Nguyen

Food Engineering Innovations Across the Food Supply Chain discusses the technology advances and innovations into industrial applications to improve supply chain sustainability and food security. The book captures the highlights of the 13th International Congress of Engineering ICEF13 under selected congress themes, including Sustainable Food Systems, Food Security, Advances in Food Process Engineering, Novel Food Processing Technologies, Food Process Systems Engineering and Modeling, among others. Edited by a team of distinguished researchers affiliated to CSIRO, this book is a valuable resource to all involved with the Food Industry and Academia.

Feeding the world’s population with safe, nutritious and affordable foods across the globe using finite resources is a challenge. The population of the world is increasing. There are two opposed sub-populations: those who are more affluent and want to decrease their caloric intake, and those who are malnourished and require more caloric and nutritional intake. For sustainable growth, an increasingly integrated systems approach across the whole supply chain is required.

• Focuses on innovation across the food supply chain beyond the traditional food engineering discipline
• Brings the integration of on-farm with food factory operations, the inclusion of Industry 4.0 sensing technologies and Internet of Things (IoT) across the food chain to reduce food wastage, water and energy inputs
• Makes a full intersection into other science domains (operations research, informatics, agriculture and agronomy, machine learning, artificial intelligence and robotics, intelligent packaging, among others)

• Language: English
• Publisher: Academic Press
• Release date: Dec 5, 2021
• ISBN: 9780323853590

Book preview

Chapter 1

Understanding and building resilience in food supply chains

R. García-Floresa, Simon Allenb, Raj Gairec, P. Johnstoned, D. Yie, J. Johnstond, T. Whitef, D. Scotlandd, Lilly Lim-Camachog

aCSIRO Data61, Docklands, VIC, Australia

bCSIRO Agriculture and Food, Hobart, Tasmania, Australia

cCSIRO Data61, Canberra, ACT, Australia

dNew Zealand Institute for Plant & Food Research, Hawke’s Bay, New Zealand

eNew Zealand Institute for Plant & Food Research, Auckland, New Zealand

fNew Zealand Institute for Plant & Food Research, Ruakura, New Zealand

gCSIRO Agriculture and Food, Brisbane, QLD, Australia

Abstract

The economic and natural environments in which fresh produce supply chains operate are continuously changing, and more so with the challenges that climate change and a growing population impose. In order to adapt to these changes, there is a pressing need to quantify and understand the risks and identify the strategies that can minimize them. The purpose of the present chapter is, first, to review existing work and ideas relating to resilience and robustness from the perspective of fresh produce supply chains, and second, to propose a quantitative framework to measure resilience and help fresh food producers understand how the structure of their supply networks may enhance their resilience when facing disruptions based on the well-known framework provided by the risk matrix. We apply the framework to a case study of apple production and distribution in New Zealand.

Keywords

Risk assessment; Resilience; Robustness; Supply chain risk management

1.1 Introduction

The economic and natural environments in which food supply chains operate are continuously changing, and more so with the challenges that climate change and a growing population impose. In order to adapt to these changes, there is a pressing need to quantify and understand the risks and identify the strategies that can minimize them. The International Standards Organization defines risk as the effect of uncertainty on objectives (ISO, 2009). Risk assessment is the systematic process of evaluating the potential risks that may be involved in a projected activity, and risk management is the identification of procedures to avoid or minimize the impact of identified risk’s impact. Risk is everywhere and its sources are mostly probabilistic variation (aleatory uncertainty) and lack of knowledge (epistemic uncertainty).

Given the importance of the topic, several reviews on supply chain risk management (SCRM) are available. Ho et al. (2015) provided an exhaustive review with the aim of covering all aspects of SCRM, including supply chain risk types, risk factors, and risk management methods. Ho et al. (2015) also proposed a conceptual framework to classify the papers they reviewed, and which we will use as a guide to demonstrate our methodology to quantify resilience (Fig. 1.1). Bak (2018) reviewed the existing literature and identified knowledge gaps and research trends, noting that most existing publications are biased to case studies in the United States and the United Kingdom. Fan and Stevenson (2018) provided another review with the aspiration of providing a new definition of SCRM and contribute to its theory with an object-process-outcome framework. Another general review was completed by Zhu et al. (2017), who discussed integrated SCRM and point to potential research directions.

Fig. 1.1 The conceptual framework of supply chain risks, adapted from Ho et al. (2015).

Reviews addressing specific aspects of SCRM are more numerous. For example, Negreiros de Oliveira et al. (2019) who, motivated by the many recent environmental scandals and accidents, analyzed environmental risk factors and provide a taxonomy and a framework that synthesized environmental risks, consequences, and strategies covered by academic research. The framework they developed is quite complete in that it explicitly considers that the consequences for businesses of environmental negligence are not only financial in the shape of fines. Companies can also face severe reputational consequences, boycotts from customers, negative media exposure, and loss of credibility that can be irreversible.

Optimization methods are an important technology that is commonly used to assess and mitigate risk in supply chains. These can be deterministic, when it is assumed that all the information is known beforehand, or stochastic, which are methods that generate and use random variables that commonly represent scenarios. These methods used to inform strategic investment and operational decisions. Optimization can be used to analyze every stage or combination of stages in the supply chain, for example, transportation and logistics (Li et al., 2020; Alkaabneh et al., 2020), location of centralized collection, and processing facilities (Domingues-Zucchi et al., 2011; Garcia-Flores et al., 2014), reduction of losses prefarm and postfarm gate (Paam et al., 2019; Banasik et al., 2017) and others. These methods are becoming crucial given the increasing constraints imposed to production by scarcer resources and social needs. The most effective way to increase the amount of food produced is through efficiency gains, as clearing new land for agricultural activities stopped being sustainable a long time ago. For example, García-Flores (2015) carried out a study of the efficiency of investment to support small communities of cheese makers to add value to their whey by-product, reduce food losses and decrease environmental damage of disposal. The supply chain was designed as the optimal configuration of the whey supply chain and demonstrated that important savings and benefits to the community could be achieved by investing early on adequate processing facilities, which also increased the resilience of the system.

We will center the discussion in this chapter mostly on the concepts of resilience and robustness in the context of agricultural supply chains. Behzadi et al. (2018) reviewed quantitative models for agribusiness SCRM with focus on these two terms, which they define as follows: robustness is an ability to withstand disruption with an acceptable loss of performance whereas resilience is the potential to recover quickly from disruption. Robustness is a suitable capacity for managing business-as-usual risks (i.e., high probability, low impact risks to be mitigated), while resilience is suitable for disruption risks (i.e., low probability, high impact risks, or contingencies). Roy (2010) defines robust as the capacity for withstanding vague approximations and/or zones of ignorance in order to prevent undesirable impacts, notably the degradation of the properties of the system that the stakeholders want to be maintained. Ben-Haim (2012) (cited by Aven, 2016) affirms that in decisions under uncertainty, what should be optimized is robustness rather than performance, or in other words, the decision maker should prefer continuity and satisficing rather than optimizing. Resilience and robustness together ensure that a system is reliable, that is, it is free from failure and can perform consistently well. A useful framework to unify some of these concepts is the risk matrix (Fig. 1.2).

Fig. 1.2 The risk space framework.

The level of resilience for a system or organization is linked to the ability to sustain or restore its basic functionality following a stressor. A resilient system can (Hollnagel et al., 2006, cited by Aven, 2016):

• respond to regular and irregular threats in a robust yet flexible (adaptive) manner,

• monitor what is going on, including its own performance,

• anticipate risk events and opportunities,

• learn from experience.

Although resilience is a generic term, it is most used in the safety domain, whereas robustness is most commonly referred to in business and operational research contexts.

A related term is vulnerability, which is defined in as the propensity of risk sources and risk drivers to outweigh risk mitigation strategies, thus causing losses and adverse supply chain consequences (Jüttner et al., 2003).

Resilience stands out as a promising research topic to address change, although it is inherently complex because it involves not only a response to disruption, but also touches on the cultural aspects of continuous improvement; these challenges are briefly discussed in the next section. Hamel and Valakingas (2003) include resilience as one of the three common forms of innovation:

1. Revolution, or creative destruction,

2. Renewal, or creative reconstruction, and

3. Resilience, or the capacity of continuous reconstruction.

In making the case for continuous reconstruction, Hamel and Valakingas (2003) note that, contrary to a very common business misconception, the idea of novelty is not fundamentally attached to risk, but that risk is a function of uncertainty multiplied by the size of the decision maker’s financial exposure; novelty defies convention, and confusing newness with risk misleads companies to continue investing in old, decaying strategies based in past successes. According to the extensive literature review of Behzadi et al. (2018), strategies to enhance resilience have been studied far less than strategies to give robustness to supply chains, a similar conclusion as Barbosa-Póvoa et al.’s (2018), who state that work exploring sustainable supply chains’ resilience is practically inexistent.

Regarding reports on case studies for making agricultural supply chains more resilient, there are abundant examples and case studies of supply chain adaptations in response to a wide range of disruptions. Fujisawa et al. (2015) studied the strategies for adaptation to climate change of apple farmers in three regions of Japan and South Africa, and found that their responses changed depending on the way their business is organized, with cooperatives being more likely to adopt a top-down approach to decision making, whereas farmers who have established their own sales channels tend to act on decisions that are born from the bottom up. Leguizamon et al. (2016) describes a program from a major retailer to buy directly from small farm suppliers, implementing a support strategy that commits to increase the resilience of the supply chain. Among Leguizamon et al.’s (2016) conclusions is that including a broad base of farmers in the supply chain is important for both suppliers’ income and the supermarket’s supply practices. Fujisawa et al. (2015) acknowledge that a combination of bottom-up and top-down leadership will facilitate more flexible and easily accepted policies for adaptation to a changing environment.

The purpose of the present chapter is, first, to review existing work and ideas relating to resilience and robustness from the perspective of fresh produce supply chains such as horticulture, grains, fish, and meat, and second, to propose a quantitative framework to measure resilience and help fresh food producers, packers or processors understand how the structure of their supply networks may enhance their resilience when facing disruptions based on the risk matrix summarized in Fig. 1.2. To that end, we discuss the risk space that supply chains are immersed in. We apply the framework to a case study of apple production and distribution in New Zealand.

1.2 The challenges for the supply chains of fresh produce

Disruptions to fresh produce supply chain operations may be produced by economic, environmental, or social factors. Identifying the most likely sources of disruption will depend on the type of supply chain under study, as each productive activity will be subject to its own specific challenges. To cite some examples, fruit and vegetable production is dominated by the effect of climate while being grown in the field; grains are very sensitive to the quality of transportation systems; the demand for beef has been increasing worldwide, but the cattle industry competes for resources such as land and water with society and other industries (García-Flores et al., 2015). To cope with disruptions and become more resilient, companies face the following internal challenges (Hamel and Valakingas, 2003):

1. The strategic challenge, which relates to medium- and long-term planning. Organizations must consider alternatives and create new options to update and replace strategies as they age and become less effective.

2. The cognitive challenge, which is a cultural concern. Companies must free themselves from denial, nostalgia, and arrogance.

3. The political challenge, which is related to resource allocation. Funding and skills must find their way to prototypes and experiments that may become tomorrow’s products and services.

4. The ideological challenge is related to the search for efficiency. Optimization alone cannot make a business model more relevant when a changing environment dictates that the current operations are not effective anymore.

The social and cultural nature of these challenges, especially the last three, demonstrates that resilience is a complex concept that involves not only the operative response to disruption, but also the cultural aspects of continuous improvement and proactive anticipation, which makes it difficult to quantify. To complicate matters, these challenges are harder to attain for networks of companies than for individual players, as more coordination and a broader view of the system are necessary. The fact that cognitive, political, and ideological agreement is needed makes the design of a truly resilient supply chain a very difficult endeavor. Resilient supply chains must be proactive, as reactive strategies present many pitfalls; an enlightening example is presented in Rice and Gavin (2006). For the sake of the argument, we will consider in the remainder of this chapter that, regardless of the nature of the disruption, the ultimate effect will be the interruption of flow caused by the removal of one or more nodes or links from a network representation of the supply chain.

1.3 Quantifying resilience

Several metrics have been proposed to quantify resilience from a strategic perspective, mostly from the fields of ecology and telecommunications. The focus in ecology has been to understand the capability of populations to recover from threats. For example, Moore et al. (2016) analyzed 9 years of ostrich movement data to explore the resilience of the Western Cape ostrich industry, which was affected after an outbreak of avian influenza in 2011 and has gradually recovered. The network metrics used to assess resilience were density, network diameter, number of weak components, size of giant component and degree distribution. They found the ostrich movement network to be resilient. Despite a decrease in the number of farms after the outbreak, the system slowly returned to its former state. Plagányi and Essington (2014) developed an index to identify key species in an ecosystem in order to measure the repercussions of strengthening or weakening fish populations on the broader ecosystem structure as well as the overall economic value of fisheries.

There is also a lot of interest in defining resilience in telecommunications, especially for cybersecurity and infrastructure risk management. Alenazi and Sterbenz (2015) performed a comprehensive comparison of the most commonly used graph robustness metrics and found the most adequate predictors of resilience for different types of networks, including some generated artificially. In the context of Internet security, Salles and Merino (2012) introduced a resilience factor based in the concept of k-connectivity and propose strategies to improve resilience by altering the network topology. Rosenkrantz et al. (2009) developed a graph-theoretic model for service-oriented networks. In this type of network, when a user requests a service that is locally available at a node, the node provides the service directly. When the requested service is nonlocal, the node forwards the user’s request to another node in the network where the service is available locally. Rosenkrantz et al. (2009) provided a resilience metric, as well as algorithms for designing resilient networks.

Regarding risk-related supply chain network metrics, Ortmann (2005) analyzed the South African fresh fruit export supply chain and modeled it as maximum flow and minimum cost flow problems, using optimization and graph theory methods. Barthélemy (2011) provided an extensive review of network metrics in the context of spatial networks, that is, networks where space is relevant and where topology alone does not contain all the information. Martin and Niemeyer (2019) discussed the implications of error measurements, which are unavoidable in networks representing real-world entities like supply chains, on the reliability of network metrics. More relevant to the methodology presented in this chapter is Lim-Camacho et al. (2015) and Lim-Camacho et al. (2017), who used a network-based simulation to estimate the resilience of supply chains and proposed four supply chain indices (SCIs) for evenness, resilience, continuity of supply, and climate resilience, to estimate the performance of fish, rice, and mineral supply chains. In particular, the resilience metric is based on the SCI introduced in Plagányi et al. (2014). This metric is described in detail in Section 4.1.

1.4 Methodology

In this Section we describe our methodology in detail, first by introducing the SCI of Plagányi et al. (2014) and then by explaining an approach to measure resilience empirically.

1.4.1 The supply chain index

The original SCI introduced in Plagányi et al. (2014) represents a measure of connectance, with lower values indicating higher connectance. In a connected network G, let N and L be the set of nodes and links, respectively, and super-index S indicate the source nodes such that NS is the set of all source nodes. Let i, j ∈ N the source and destination of a product, respectively, |L| the total number of links and |N| the total number of vertices. Let fij the flow from i to j.

Input.

1. The elements of matrix Sij, which represents the proportion of flows into each node respect to the total input flow,

(1.1)

that is, Sij is the flow from i to j divided by the total flow into node j from every supplier, such that for all nodes that are not sources in G.

2. The elements of matrix pij, which is the proportion of the total product in the supply chain that flows into receiver j,

(1.2)

that is, the flow from i to j divided by the total production of the sources, such that for all nodes that are sources in G. The element pij represents the proportion of the total product in the supply chain that flows into receiver j.

Output.

1. The SCI for each element j is

(1.3)

The critical elements in G are identified as those with the highest SCI score.

2. The supply chain index total (SCIT) for the supply chain as a whole is obtained by summing over individual