Identifying fruit and vegetable losses and waste causing factors in supply chain towards achieving sustainable consumption and production

Abstract

With a growing population and rising demand, food loss and wastage (FLW) has emerged as one of the major issues of the early twenty-first century. As per the United Nation approximately 13% of the food gets wasted after harvesting and 17% goes wasted in consumption level thus an effective food waste management is required to achieve sustainable consumption and production. The effective waste management are reliable on how the loss and waste affecting factors are controlled. Extensive studies have been conducted to determine FLW in the farming and food processing industries for both pre-and-post harvest, however, to the best of our knowledge, there hasn't been any writing on the fruit and vegetable (F&V) supply chain. from firming to retail. This research specifically examines, FLW issues in the F&V supply chain (from farm production to retail). Through a three-stage combined multicriteria decision making (MCDM) process, FLW factors are analyzed and critical factors are discovered based on influential weights and their rank. Integrated analytical network method (ANP) based on grey-decision making, trial assessment, and laboratory (DEMATEL)is used. Fifteen elements of FLW in the F&V supply chain have been identified through the literatures and opinions from industry experts. According to the grey-DEMATEL framework, eight factors are classified as effects and seven as causes. Climate change has the highest weightage of 0.099 in the ANP technique, whereas competitive market location has the lowest weightage of 0.036. Independent variables like soil fertility, cultural and traditional barriers, and bacterial infection are in the top left corner, while dependent ones like a lack of collaboration, a distant market, a lack of food policy, and not enough storage are in the bottom right. This research provides important insight for policymakers and practitioners by illuminating the role of FLW to minimise food waste policy and building proactive strategy that help towards achieving sustainable consumption and production.

Appendix: Methodology of grey-DEMATEL and ANP

1.1 Conversion of average grey relational matrix to normalized grey relational matrix

The converted grey initial direct relational matrix was normalized according to Eq. 1.

Lower and upper normalized values.

$$ \begin{gathered} \underline { \otimes } \dot{x}_{ij} = \left( {\underline { \otimes } \check{x}_{ij} - {}_{j}^{min} \underline { \otimes } \check{x}_{ij} } \right)/\Delta_{min}^{max} , \hfill \\ \overline{ \otimes }\dot{x}_{ij} = \left( {\overline{ \otimes }\check{x}_{ij} - {}_{j}^{min} \overline{ \otimes }\check{x}_{ij} } \right)/\Delta_{min}^{max} \hfill \\ \end{gathered} $$

(1)

where ${\underline{\otimes }\dot{x}}_{ij}$ implies the lower normalized limit value of the grey value $\underline { \otimes } \check{x}_{ij}$ and $\underline{\otimes }{\dot{x}}_{ij}$ implies the upper normalized limit value of the grey value $\underline { \otimes } \check{x}_{ij}$.

$$ \Delta_{min }^{max} = jmax\underline { \otimes } \check{x}_{ij} - jmin\underline { \otimes } \check{x}_{ij} $$

(2)

Compute the total normalized crisp value

$${C}_{ij} = \left(\frac{\left({\underline{\otimes }\dot{X}}_{ij}(1-{\underline{\otimes }\dot{X}}_{ij}\right)+(\underline{\otimes }{\dot{X}}_{ij}\times \underline{\otimes }{\dot{X}}_{ij})}{\left(1-{\underline{\otimes }\dot{X}}_{ij}+\underline{\otimes }{\dot{X}}_{ij}\right)}\right)$$

(3)

Compute final crisp values

$${C}_{ij}^{*}=\left(min{\underline{\otimes }\dot{x}}_{ij}+({C}_{ij}\times {\Delta }_{min}^{max})\right)$$

And

$$C=\left[{C}_{ij}^{*}\right]$$

Normalized direct relation matrix N is determined by

$$N=P*M$$

(4)

where $P=\frac{1}{{\text{max}}1\le i\le I\sum_{j}^{I}{x}_{ij}}$

For obtaining prominence and relational value, row R and column D of the total influential matrix were calculated for all the challenges. R and D for all challenges were calculated by using Eq. 6.

$$ R = \sum\limits_{j = 1}^{n} t_{ij} ,\;{\text{and}}\;D = \sum\limits_{i = 1}^{n} t_{ij} $$

(5)

For getting total influential matrix ‘T’ Eq. 6 is used which is obtained by using the normalized crisp matrix.

$$T=N\times {(I-N)}^{-1}$$

(6)

where I is Identity matrix.

1.2 Computation of normalized supermatrix and unweighted supermatrix

For the input of ANP, the initial unweighted super-matrix ‘W_x’ was obtained directly from the total influential matrix (the result of grey DEMATEL). To obtain the unweighted super-matrix ‘Wu’, the super-matrix ‘Wx’ is normalised by using Eq. 7 and obtained ‘W_n’. By transposing the super-matrix ‘W_n’, the unweighted super-matrix ‘Wu’ was obtained (Tables 8, 9 and 10).

$$ \begin{aligned} W_{x} & = \left[ {\begin{array}{*{20}c} {x_{D}^{11} } & \cdots & {x_{D}^{1j} } & \cdots & {x_{D}^{1n} } \\ \vdots & \ddots & \vdots & \ddots & \vdots \\ {x_{D}^{i1} } & \cdots & {x_{D}^{ij} } & \cdots & {x_{D}^{in} } \\ \vdots & \ddots & \vdots & \ddots & \vdots \\ {x_{D}^{n1} } & \cdots & {x_{D}^{nj} } & \cdots & {x_{D}^{nn} } \\ \end{array} } \right] \\ W_{n} & = \left[ {\begin{array}{*{20}c} {\frac{{x_{D}^{11} }}{d1}} & \cdots & {\frac{{x_{D}^{1j} }}{d1}} & \cdots & {\frac{{x_{D}^{1n} }}{d1}} \\ \vdots & \ddots & \vdots & \ddots & \vdots \\ {\frac{{x_{D}^{i1} }}{di}} & \cdots & {\frac{{x_{D}^{ij} }}{di}} & \cdots & {\frac{{x_{D}^{in} }}{di}} \\ \vdots & \ddots & \vdots & \ddots & \vdots \\ {\frac{{x_{D}^{n1} }}{dn}} & \cdots & {\frac{{x_{D}^{nj} }}{dn}} & \cdots & {\frac{{x_{D}^{nn} }}{dn}} \\ \end{array} } \right] \\ \end{aligned} $$

(7)

where ${d}_{i}={\sum }_{j=1}^{n}{x}_{d}^{ij}$; ${W}_{u}= {\left[{W}_{n}\right]}^{T}$