*Article* **Synergetic Benefits for a Pig Farm and Local Bioeconomy Development from Extended Green Biorefinery Value Chains**

**James Gaffey 1,2,3,\*, Cathal O'Donovan 4, Declan Murphy 5, Tracey O'Connor 1,2, David Walsh 6, Luis Alejandro Vergara 2, Kwame Donkor 7, Lalitha Gottumukkala 7, Sybrandus Koopmans 8, Enda Buckley 9, Kevin O'Connor <sup>2</sup> and Johan P. M. Sanders <sup>8</sup>**

	- <sup>4</sup> Carhue Piggeries, Cooligboy, Timoleague, Co., P72 HD61 Cork, Ireland
	- <sup>5</sup> Makeway Nutrition, Unit 6, Riverstown Business Park, Tramore, Co., X91 TRF9 Waterford, Ireland

**Abstract:** As the global population rises, agriculture and industry are under increasing pressure to become more sustainable in meeting this growing demand, while minimizing impacts on global emissions, land use change, and biodiversity. The development of efficient and symbiotic local bioeconomies can help to respond to this challenge by using land, resources, and side streams in efficient ways tailored to the needs of different regions. Green biorefineries offer a unique opportunity for regions with abundant grasslands to use this primary resource more sustainably, providing feed for cows, while also generating feed for monogastric animals, along with the co-production of biomaterials and energy. The current study investigates the impact of a green biorefinery coproduct, leaf protein concentrate (LPC), for input to a pig farm, assessing its impact on pig diets, and the extended impact on the bioenergy performance of the pig farm. The study found that LPC replaced soya bean meal at a 50% displacement rate, with pigs showing positive performance in intake and weight gain. Based on laboratory analysis, the resulting pig slurry demonstrated a higher biogas content and 26% higher biomethane potential compared with the control slurry. The findings demonstrate some of the local synergies between agricultural sectors that can be achieved through extended green biorefinery development, and the benefits for local bioeconomy actors.

**Keywords:** bioeconomy; biorefinery; pigs; soya bean; grass; protein; biogas; biomethane

#### **1. Introduction**

The world currently faces a climate and biodiversity emergency brought about, in part, by an unsustainable food system, which has an immense environmental footprint [1,2]. Globally, livestock farming is the single greatest source of environmental impacts in agriculture, accounting for 77% of all land used for food production, with 14.5% of global greenhouse gases (GHGs) linked to livestock production (rising to 25% if land use change is included) [3]. In Europe, agriculture is a significant contributor to GHGs with about 10% of all Europe's GHGs generated by agriculture [4]. This varies by country, with countries at the lower end, such as Netherlands, U.K., and Greece, below 10%, and other countries such as Lithuania and Latvia recording over 20% of their emissions from agriculture [5]. Ireland has the highest proportion of national agricultural emissions of all EU member states, with

**Citation:** Gaffey, J.; O'Donovan, C.; Murphy, D.; O'Connor, T.; Walsh, D.; Vergara, L.A.; Donkor, K.; Gottumukkala, L.; Koopmans, S.; Buckley, E.; et al. Synergetic Benefits for a Pig Farm and Local Bioeconomy Development from Extended Green Biorefinery Value Chains. *Sustainability* **2023**, *15*, 8692. https://doi.org/10.3390/su15118692

Academic Editors: Idiano D'Adamo and Massimo Gastaldi

Received: 28 February 2023 Revised: 21 May 2023 Accepted: 22 May 2023 Published: 27 May 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

over 30% of total national emissions coming from agriculture [5,6], the vast majority of which arises from livestock sectors [6].

The European Green Deal sets out Europe's pathway towards a climate-neutral future, which involves reducing net emissions by at least 55% by 2030 and ultimately achieving climate neutrality by 2050 in line with its commitments under the Paris Climate Agreement [7]. These ambitious European targets place pressure on the agriculture sector to reduce its emissions at the member-state level, with several member states already taking sector-specific action in response.

The development of a circular bioeconomy underpinned by sustainable renewable biological resources from primary production and its associated technological, product, and systemic solutions has been proposed as offering a solution to support a reduction in emissions within the broader agricultural and livestock sectors, while at the same time offering new business and innovation opportunities in traditional primary sectors [8]. Solutions vary from synthetic and seaweed-based additives, which can potentially inhibit rumen methane emissions [9,10], to the displacement of fossil-based materials and fuels with new bio-based materials and biofuels produced from agricultural by-products [11–13], to nature-based solutions [14,15]. Many synergies for bioeconomy development exist within primary sectors, and some of these are explored in Figure 1. Using this approach, many local agricultural and societal needs may be met using available local primary resources, thereby increasing the self-sufficiency of the region, while reducing emissions by using local and shorter supply chains. This approach may be seen as an expansion of the concept of bio-districts or eco-regions which, as described by Dias et al. [16], aims to stimulate a collective approach to sustainable resource management by adopting organic farming practices at a territorial level to ensure benefits are distributed across the region, stimulating rural stakeholder networks and strengthening existing links between rural stakeholders. This can contribute to sustainable local development that prioritizes resource conservation and ecological integrity as inherent characteristics of economic logic, and can contribute to greater quality of life for direct stakeholders and local communities [16]. The potential of adapting this bio-district approach to support local bioeconomy development is exemplified in Figure 1 below, which shows potential synergies between (1.) grassland, (2.) monogastrics, and (3.) marine sectors, which can all provide goods and services for the overall bio-district.

One opportunity for potential local symbiosis between primary sectors exists in the development of green biorefinery models. Many protein sources are not being used optimally and may be more efficiently used by deploying biorefinery approaches [17]. In green biorefinery systems, green biomasses, such as grasses and legumes, can be processed into multiple co-products, including leaf protein concentrate (LPC), which is suitable for both ruminant and monogastric animals and for use in aquaculture [18–20]. This approach can help to improve the overall resource efficiency of the green biomass by providing an ensiled fiber press cake for cows, which can offer a reduction in nitrogen and phosphorous emissions while maintaining milk productivity [18,20,21], and by providing an additional LPC co-product that can potentially replace soya bean meal in pigs, poultry, and other monogastrics [22,23]. The opportunity to use available grasslands in Europe more efficiently to produce a local, homegrown protein could resolve a key challenge for European agriculture, which is heavily reliant on imported sources of unsustainable feed. The European Union (EU) is reliant on animal feed imports from North and South America for livestock production, due to domestic deficits [24]. In addition, such a model can provide further bio-materials including fiber for insulation materials, bio-composites and packaging, high-value prebiotic materials, and brown juice or grass whey [25,26]. This whey can subsequently be used in different applications, for example, in biogas production or as fertilizer [25,26]. The interest in biogas production and its upgrading to biomethane, which can substitute natural gas and vehicle fuel, has become a trend in Europe, but its potential is still underexploited [27–29]. This interest has been further heightened by the Russia–Ukraine war. The REPower EU Plan, introduced by the EU Commission in 2022,

focuses on reducing energy dependence on Russia by setting a target of 35 billion cubic meters of biomethane production by 2030, thus replacing the need for the import of natural gas [30].

**Figure 1.** A bio-district approach for bioeconomy development. The figure shows potential synergies between (1.) grassland (ruminants), (2.) monogastrics, and (3.) marine sectors, which can collectively provide goods and services for the overall local bio-district (4.). Conventional products (traditional food products, such as milk, beef, fish, and pork) are highlighted by continuous lines while dotted lines indicate new value chains and products. A green biorefinery (5.) and anaerobic digestion unit (6.) are included as examples of enabling technologies that can support new local value chain development. In sector (1.), grass is supplied into a green biorefinery which can create low-emission feeds for ruminants and monogastrics (and potentially aquaculture feed), replacing imported feed sources. Other bio-material products may also be produced such as fiber insulation materials. Conventional dairy and meat products are also produced from the grassland chain. In sector (2.), pig slurry along with other animal slurries and food waste from the municipality can be supplied to an anaerobic digester to create heat, electricity, and/or biofuels for the bio-district. Pork is also produced. In sector (3.), coastal/marine resources can supply food products such as fish, while processed marine biomass, such as microalgae and Asparagopsis, can be used as fertilizer, feed, or anti-methanogenic additives for ruminants within the grassland chain.

This article explores the potential benefits of green biorefinery co-product LPC for a pig farm, partially displacing imported soya bean meal within a conventional pig diet. The impact of this diet change on pig performance is considered, as is the resulting pig slurry including its impact on the biomethane potential of slurry, a key factor for the pig farm which supplies its slurry to a local community anaerobic digestion plant. While a previous study has investigated the potential of green biorefinery LPC as an alternative pig feed [31], within this current study the inclusion rate of LPC was greatly increased to understand the impact of a higher inclusion rate. Furthermore, while other studies have investigated the biogas and biomethane potential of green biorefinery by-products such as grass whey, brown juice, de-FOS whey [25,26], and press cake [25], and various studies have looked at pig slurry [32–34], no similar studies investigating and analyzing the impact of the resulting biorefinery LPC pig slurry from this value chain were found within the

literature This represents a novel aspect of the current research. In this way, this paper explores, in more detail, the potential benefits that green biorefineries implemented within the cattle sector can deliver for the pig sector.

#### **2. Materials and Methods**

Fresh grass was harvested from farms located within a 10 km radius of Afferden, Limburg, in the Netherlands during July and August of 2021. The feedstock, a 75–25% perennial ryegrass–clover mix, was processed within 12 h. The processing was carried out with the innovative "green biorefinery" process developed by Grassa BV. The biorefinery is a fixed unit demonstration facility with 4 tonnes per hour of fresh input processing capacity. A schematic of the process is included as part of Figure 2 below. Briefly, fresh grass was washed with water upon entry to the biorefinery to remove sand, with the water being recycled within the process. The grass was then subject to wet fractionation using an extrusion process. This created two primary products, a protein-rich "press cake" containing 50–60% of the initial grass protein, and a green, grass-derived juice. The press cake was further preserved through ensiling and baling, to be later used as ruminant feed, which exhibits a high nitrogen use efficiency and reduced nitrogen and phosphorous emissions [20]. The remaining protein is contained within the green juice. The protein fraction of the green juice was solidified using a heat coagulation process. This solid protein portion was extracted using a vacuum filter and then dried to in excess of 90% dry matter (DM) creating a storable LPC. The LPC was incorporated into appropriate feed for the pigs involved in this study. The remaining grass whey retains valuable sugars, e.g., fructans, and minerals after the protein fraction has been removed, and these can be extracted through further processing. The Grassa green biorefinery process is presented in Figure 3a below. The LPC was later tested as a feed ingredient at Carhue Piggeries in Timoleague, Co., Cork, Ireland. The piggery is also linked with Timoleague AgriGen Biodigester, where it currently supplies its generated pig slurry for the purposes of biogas production (Figure 3b below). The green biorefinery process is highlighted in Figure 2 below, with the main focus points of the current study, testing of LPC input to the pig farm and impacts on pig and biogas production, highlighted by a dotted line.

**Figure 2.** Schematic of green biorefinery value chain with main focus areas of the current study highlighted with a dotted line.

$$\mathbf{(a)}\tag{b}$$

**Figure 3.** (**a**) Grassa 4-tonne per hour pilot green biorefinery plant production unit for LPC; (**b**) Timoleauge AgriGen Biodigester for supply of slurry to produce biogas.

#### *2.1. Analysis of Green Biorefinery LPC as an Alternative Feed for Pigs*

#### 2.1.1. Characterization of LPC

A proximate analysis of LPC was performed by Dairygold's analytical laboratory at Mallow, Co., Cork, Ireland. Crude fiber was analyzed using the Fibretec method based on ISO 6865:2000 [35]. Protein was determined using the Semi-Micro-Kjeldahl method based on ISO 5983:2000 [36]. Calcium, magnesium, sodium, and potassium were determined using atomic absorption based on ISO 6869:2000 [37]. Moisture content was determined using an oven at 102◦C based on ISO 6496:1999 [38]. Phosphorous was determined using colorimetric UV spectrophotometry following ISO 6491:1998 [39]. Starch was determined using the polarimetry method based on ISO 15914:2004 [40]. Oil content was determined using the Soxhlet method following ISO 6492:1999 [41]. Crude ash was determined using a muffle furnace based on ISO 5984:2002 [42].

The LPC was analyzed by Sciantec, North Yorkshire, U.K., to assess the amino acid profile. Cation exchange chromatography was used to measure the AA content, following 24 h of exposure of the LPC to acid hydrolysis.

#### 2.1.2. LPC and Control Diet Pig Feed Trial

To explore the potential of the LPC biorefinery co-product as a partial substitute for soya bean meal in pig diets, pig feeding trials were conducted at a privately owned piggery, Carhue Piggeries at Timoleague, Co., Cork, Ireland. Pigs were fed from two separate silo's via Schauer's Spotmix feeding system, in which feed is dispensed dry and is combined with water at the valve above the feeding trough and is presented to the pigs as wet feed. This system facilitates the feeding of multiple diets across pen groups without cross contamination or feed freshness issues. Pigs were fed ad libitum with feed troughs probed by the computer multiple times per hour with the same probe preventing over-feeding by regulation by the same feeding probe. Based on the characterization of the LPC, treatment and control feed rations were designed by Makeway Nutrition Ltd. with input from contributors with a view to reducing the soya bean usage of a conventional pig weaner ration by 50%. The nutritional requirements of the weaner pigs were considered in the preparation of the treatment diet in order to make it comparable to the control diet and sufficient for the pigs' needs. The treatment diet is provided in Table 1 alongside the control diet.


**Table 1.** Treatment and Control diet formulations for weaner pigs.

The focus of the trial was on weaner pigs aged approximately 9 weeks old and weighing 17 kg on average at the start of the testing phase. The pigs were randomly split into a treatment group of 110 pigs to be fed the treatment feed and a control group with 110 pigs to be fed the control diet over a 31-day period. The control diet was comprised of wheat, maize, barley, molasses, soya bean meal (in the form of hipro soya), soy oil, soy hull, and minerals in recommended amounts (Table 2). The treatment feed included LPC replacing 50% of the hipro soya level contained in the control diet, which represented the most significant change between the two diets (Table 2). Small amounts of synthetic lysine and methionine (0.08% and 0.04% of diet, respectively) were added to the LPC diet to prevent the crude protein of the diet from exceeding 18% crude protein and differing greatly from the control diet. All other amino acids are non-limiting. Additionally, a small amount of limestone flour was removed from the grass protein diet due to the high calcium value within the LPC. A comparison of constituents from the control and treatment diet is displayed in Table 2. The two diets were dispensed from separate feed silos using a computerized feed system which dispensed weighed amounts of feed to each group of pigs.

**Table 2.** Main constituents of treatment and control diets.


The weight of pigs and feed intake were recorded on a weekly basis. These data were necessary in order to calculate the feed intake and weight gain on a daily basis (daily feed intake and average daily gain), and to evaluate the feed conversion ratio per treatment.

The daily feed intake (DFI) [43] for each treatment was evaluated by dividing the total feed intake by the size of the group, i.e., the number of pigs. The total feed intake was calculated at the end of the trial by deducting the feed remaining from the amount of feed delivered.

> Daily Feed Intake <sup>=</sup> Total Feed Intake Number of pigs on treatment

The pigs were weighed throughout the trial, at the start and end of every week, to calculate the average daily gain (ADG).

To calculate the feed conversion ratio (FCR) [44], the DFI was divided by the ADG for each treatment group.

$$\text{Feed Concentration Ratio} = \frac{\text{Daily Feed Intake}}{\text{Average Daily Weight Gain}}$$

#### *2.2. LPC and Control Slurry Biogas and Biomethane Analysis*

To assess the potential effect of the diet change on the slurry feedstock from the perspective of anaerobic digestion, a biogas and biomethane analysis of the feedstock was undertaken. Carhue Piggeries currently sends its produced pig slurry to Timoleague Agri Gen, a community-based anaerobic digestion facility located at Timoleague Co., Cork, Ireland approximately 2 km from the piggery site. At this facility, the slurry is co-digested along with food waste to produce biogas which is converted to heat and electricity. The site produces 500 KW of renewable electricity. The digestate produced from the digestion process is rich in nitrogen (N), phosphate (P), and potassium (K), and this biofertilizer can be land spread as a substitute for chemical fertilizer. To complete the analysis, approximately 2 kg of mixed slurry resulting from the treatment and control pig feeding trial batches from Carhue piggeries were collected and sent for analysis by Celignis Analytical, located at Castletroy, Co., Limerick, Ireland.

#### 2.2.1. Determination of Total and Volatile Solids for Slurries

Slurries from weaner pigs feeding on LPC and the traditional soya bean meal diets were analyzed for biomethane potential. The samples were extracted simultaneously from the manure cellars when the feeding period ended. Prior to the BMP analysis, both slurries underwent a proximate analysis for determination of the total solids, volatile solids, and ash content. The proximate analysis followed European standard protocols as described in reference methods EN 14774-1:2009 [45] and EN 14775:2009 [46].

#### 2.2.2. Biomethane Potential (BMP) of Slurries

An Anaero BMP unit consisting of 15 slots for 1 L digesters (700 mL working volume) was utilized for determination of methane potential from the slurries. The digesters were constantly stirred with stainless steel paddle systems and were placed in a 37 ◦C water bath. The BMP analysis followed the German standard method protocols (VDI 4630). Celignis propriety inoculum, which treats different sources of substrates including grass, whey, sewage sludge, manure, and other lignocellulosic biomass, was utilized for the BMP analysis. The inoculum was sieved to remove residual organic matter and degassed for at least seven days to remove any remaining organic matter. An inoculum to substrate ratio of 4:1 was prescribed in the VDI 4630 reference method. The BMP analysis of each slurry was performed in triplicate with no pH adjustments. pH adjustments with special reagents were not required because the pH of the substrate–inoculum mix was within the optimum range for anaerobic digestion. The BMP analysis of LPC and soya bean meal pig weaner slurries was monitored within 28 days of digestion.

#### 2.2.3. Biogas Analysis and Biomethane Calculations

In the Anaero BMP unit, a tipping bucket flowmeter system coupled with a recording computer was used to measure biogas production from the various digesters. The flow meter had 15 chambers (buoyancy bucket design in every chamber), and each chamber contained a salt solution that prevents the dissolving of carbon dioxide, hydrogen sulfide, and ammonia from the biogas generated during this study. The biogas from the flowmeter was collected in 2 L Tedlar bags and methane, carbon dioxide, hydrogen sulfide, ammonia, and oxygen were analyzed using a Biogas 5000 gas analyzer. On the 3rd, 7th, 14th, 21st, and 28th days of digestion, the gas composition was analyzed and utilized to determine the biomethane potential of the slurries, applying methods used by Ravindran et al. [25] to evaluate the biomethane potential of green biorefinery products (Equations (1)–(5) [47–51]).

$$\text{Cumulative biomass produced} = \sum\_{\text{Day 0}}^{\text{Day 28}} \left( \text{biogas produced per day} (\text{mL}) \right) \tag{1}$$

$$\text{Biogas yield} = \frac{\text{Cumulative bioavailability (mL)}}{\text{FM or TS or VS fed (g)}} \tag{2}$$

Average CH4 percentage = Determined from Biogas analyzer 5000 (3)

$$\text{Cumulative CH}\_4 \text{ produced} = \sum\_{\text{Dry 0}}^{\text{Dry 28}} (\% \text{ Average CH}\_4 \times \text{biogas produced per day} (\text{mL})) \tag{4}$$

$$\text{Methane yield} = \frac{\text{Cumulative methane produced (L)}}{\text{FM or TS or VS fed (kg)}}\tag{5}$$

#### **3. Results**

*3.1. LPC and Control Diet Pig Feed Trial Results*

3.1.1. Characterization of Grass-Derived LPC

A number of important LPC properties including crude fiber, ash, protein, starch, and total solids were determined through proximate analysis. Based on the analysis, the LPC had a high DM content, containing only 5.5% moisture. The sample contained 42.8% crude protein on a DM basis, 3.9% crude fiber, 9.4% ash, 2.8% potassium, 0.37% phosphorous, and oil content of 12.1%. Neutral detergent fiber was 43.5%, and acid detergent fiber was 7.63%.

#### 3.1.2. Amino Acid Profile of LPC

The amino acid profile of the LPC is provided in Table 3. The analysis shows that the LPC was rich in glutamic acid (3.7%), aspartic acid (3.53%), leucine (3.10%), and alanine (2.24%). Table 4 provides a comparison between the constituents of LPC compared with soya bean meal and other common feedstuffs. Overall, the LPC compares quite well with soya bean meal, providing comparable levels of protein, lysine, methionine, and threonine, although with lower cystine. The LPC also compares favorably with rapeseed meal, with LPC containing higher crude protein and crude fiber contents than rapeseed meal (Table 4). Comparing the composition of the LPC with the results of LPC produced by Ravindran et al. [31] from perennial rye grass, the overall CP is significantly higher (43% versus 34%) and more consistent with the CP content of soya bean meal, as indicated in Table 4. The sample also has a higher content of production-limiting amino acids lysine, methionine, and threonine. The increase in protein content may be partly resulting from the removal of small fibers from the protein products and soluble salts such as potassium, which are washed using a vacuum filter in the current biorefinery process. The DM content of the LPC from this study is higher when compared with Ravindran et al. [31] (94.5% versus 87%) due to a new drying process that was implemented in the current biorefinery process. The protein was dried at 70 ◦C in a belt dryer with a residence time of 24 h; Maillard reactions occurred, as well as polymerization reactions in unsaturated fats. The crude fiber of the current LPC is lower than that found by Ravindran et al. [31], but it is still in the comparable range with soya bean meal (Table 4) [33]. The ash content was reduced from 11.9% to 9%. This reduction may be attributed to lower sand contained on the feedstock or improved washing of feedstock during pre-processing.


**Table 3.** Amino acid profile of LPC.

**Table 4.** Crude fiber, crude protein, and selected amino acid profile of various feedstuffs in comparison with LPC from the current study [52] (g/100 g).


Once mixed into rations at the Barryroe feed mill, finished feed appeared dark green in color compared with the control sample (Figure 4 below).

**Figure 4.** Control feed (**left**) compared with treatment feed (**right**)—note the green color of the treatment feed.

#### 3.1.3. Feeding Trial

Pig weights were recorded at the beginning of the trial. The initial average weight of the control group pigs was 17.388 kg, while the initial average weight for the treatment group pigs was 17.246 kg. As expected, feed intake increased in both groups during the course of the trial, and by week 5 the average trial end weight was 35.092 for the control diet and 35.450 for the treatment diet. The total weight gain per pig since weaning is presented in kg in Table 5and shows a comparable weight gain for pigs on the control diet versus those on the treatment diet, with pigs gaining 19.494 kg and 19.554 kg, respectively, by week 5.


**Table 5.** Total weight gain per pig since weaning on treatment and control diet.

The dung consistency appeared similar in both batches,; however the treatment dung had a notable green color. Overall, the pigs on the treatment diet appeared healthy.

#### Daily Feed Intake

The DFI was measured weekly during the trial. The results indicate that the pigs readily accepted the treatment feed. During the first week, week 1, the control group DFI was 1.060 kg/d and the treatment group DFI was similar: 1.079 kg/d. During week 2, the DFI dipped lower (1.155 kg/d) for the treatment diet than the control diet (1.167 kg/d). Subsequently, the DFI of the treatment diet overtook that of the control diet from week 3 onwards and was 3% higher at the end of the trial (1.427 kg/d versus 1.391 kg/d, respectively, for treatment and control).

Overall, the difference in DFI between the treatment and control group was not significant. A slightly larger amount of the treatment feed was eaten by the pigs in the treatment group than the amount of control feed eaten by their counterparts in the control group, which suggests that the weaner pigs liked the treatment diet. These results are described in Table 6. A similar trend was also seen in the study by Ravindran et al. [31], with pigs also consuming more treatment feed [31]. These findings indicate that the incorporation of green protein in the diet enhances the attractiveness of the feed, e.g., due to changes in taste, smell, or other sensory characteristics. Stødkilde et al. [23] found that the addition of green protein to pig feed rations did not negatively change the taste, i.e., pigs were not discouraged from consuming it [23]. Moreover, inclusion of green protein from specific feedstocks, e.g., clover grass, has also been found to improve the meat yield and omega-3 fatty acid content of pig meat, which may have positive health benefits [24].

**Table 6.** Comparison of treatment and control groups describing daily feed intake, feed conversion efficiency, and average daily gain.


Average Daily Weight Gain

Average daily gain (ADG) is an evaluation of the average daily increase in the live weight of an animal and is recorded across the growth period of the animal. This can provide insight into the animal's growth rate, and evaluate the time at which it should reach market weight [53]. Ravindran et al. (2021) noted a number of factors that can influence pig weight, including genetic differences, sex effects, weight at birth, age at weaning, feeding level, and specific amino acid content (e.g., arginine) in feed [31].

During the feed trials, pig weight was measured individually, first at the beginning of the trial and at the end of each following week. In week 1, pigs in the control group gained 0.630 kg/day, increasing to 0.656 kg/day after week 2, 0.675 kg/day after week 3, 0.632 kg/day after week 4, and 0.650 after the final week. For pigs on the treatment diet, the ADG started well at 0.665 kg/day at the end of week 1, but reduced to 0.611 kg/day by week 2. From that point, the ADG increased to 0.624 kg/day by end of week 3, 0.650 kg/day by end of week 4, and 0.652 kg/day by the end of the trial, slightly higher than the control diet. Overall, the variances in average weight gain between the weeks is considered to be negligible and are generally consistent across both diets. The results are presented in Table 6. Over the course of the trial, the analysis showed comparable performance in weight gain for pigs on the control and treatment diets, with pigs gaining 19.49 kg in the case of the control diet and 19.55 kg in the treatment diet by the end of the trial. The ADG of the treatment group pigs was marginally higher than the control group pigs during week 1, but dipped in the following week as the control ADG increased. This may have been a result of acclimatization of the pig's microbiome to the new protein diet. Over the course of the remaining weeks, the ADG increased in the treatment batch week on week, culminating at week 5, at which point the treatment pigs had a slightly higher (0.652 kg/day) ADG compared with the control batch (0.650 kg/day). Over the course of the trial, the ADG between both trials were quite evenly matched and without significant differences. While there were some variances in ADG week on week, these were relatively minor and did not indicate a major trend. Certain factors such as time of day with weighing or gut fill can have a big impact on the weekly weighings during trials; however, over an extended period these factors do not contribute the same variance.

#### Feed Conversion Ratio

The feed conversion ratio (FCR) describes the quantity of feed required for pig weight to increase by one pound. Lower FCR values indicate that pigs are efficiently converting feed into body weight, while a high FCR can be a sign of the pigs being unable to convert the feed into body weight effectively [31]. A number of factors influence the feed conversion ratio in pigs. Pierozan et al. (2020) found the number of pigs per pen, feeder type, origin, and sex of the animals to be determinant factors [54]. According to Hancock and Behnke (2000), feed conversion efficiency, and thus FCR, can be enhanced by pelleting feed as pigs will not sort and waste feed pellets [55]. Smaller pellets are also more digestible [55]. Lastly, dietary components such as lysine and phosphorous levels are important. Lysine is essential for pigs in order to effectively utilize other amino acids for growth [56]. Additionally, phosphorus at optimum levels is required for the proper development of muscle and optimal energy use, with excess phosphorus being excreted as feces [57].

The FCR increased weekly for both groups over the course of the trial. Initially, the FCR was recorded in week 1 as 1.68 for the control group, and 1.62 for the treatment group, which was slightly lower. In subsequent weeks, the FCR increase for both diets was quite comparable, with the FCR by the end of the trial being slightly lower for the control diet (2.14) compared with the treatment diet (2.19). The comparable FCR during the trial indicates that the treatment diet compares well to the conventional diet fed to the pigs. The comparative FCR by the end of the trial indicates a slight improvement in FCR compared with previous research on LPC [31]. The difference that exists may indicate that the regression equations used for energy estimation of the grass protein may need to be refined slightly. However, once again, the difference between the groups is minor and the performance is largely in line with the control diet. This is a field of primary concern to pig farmers as optimizing FCR can have a significant positive effect on overall profitability.

#### *3.2. LPC and Control Slurry Biogas and Biomethane Analysis Results*

The biomethane potential is a measure of the biomethane that can extracted from an organic substrate. Pig slurry is a well-recognized source of animal manure for biogas production when co-fed with carbohydrate-rich feedstock. Pig slurries have high biomethane potential (275–450 L/Kg VSfed) when compared with other animal manures except for poultry litter (460 L/Kg VSfed) [58].

The biogas and biomethane production profiles from both LPC and soya bean meal pig weaner slurries are shown in Figure 5. The maximum biogas and biomethane yields for both slurries was achieved after 25 days of digestion. The biogas production from both LPC and soya bean meal pig weaner slurries were determined to be 495 ± 12.52 L/Kg VSfed and 478 ± 7.93 L/Kg VSfed, respectively (Table 7). An ANOVA analysis conducted on the biogas yield showed no statistical significance (*p*-value (0.1206) > 0.05). This indicated no significant difference in biogas production from both LPC and soya bean meal pig weaner slurries. From Figure 5, it can be observed that 90% of biogas production was achieved after 10 to 11 days. The daily biogas production was similar for both LPC and soya bean slurries which recorded a high of 66 L/Kg VSfed. This aligns with the findings of Santos et al. [59] and Miroshnichenko et al. [60], which indicated high daily biogas production in anaerobic digestion of some pig slurries [59,60]. However, there was a significant difference in the daily methane production, which was between 10% to 50% higher in LPC slurry than soya bean meal slurry. This was attributed to the high methane content of produced biogas from the LPC, which was about 22% to 35% higher than soya bean meal slurry (statistical significance at *p*-value 0.0335 < 0.05) (Table 7). Biogas from anaerobic digestion of LPC slurry had methane contents ranging from 70% to 73% compared with the soya bean meal slurry which had biogas methane contents from 52% to 59%. The high methane content obtained for LPC could be attributed to the higher volatile solid content of the LPC pig slurry, especially with regard to the high-protein diet fed to the pigs. The significant difference in biogas methane content for both slurries led to a higher biomethane potential of 355 ± 9.45 L/Kg VSfed for LPC pig weaner slurry compared with 281 ± 7.11 L/Kg VSfed for soya bean meal pig weaner slurry (statistical significance at *p*-value 0.0004 < 0.05). The BMP results indicated that slurry from weaners feeding on the LPC treatment diet performed significantly better (26% increase in methane yield) than the conventional slurry from weaners on soya bean meal. This is potentially a major benefit in addition to the successful replacement of the soya bean mealdiet with the LPC treatment diet and suggests that co-digestion with carbohydrate-rich substrates at the Timoleague, Co., Cork, community anaerobic digester has the potential to yield improved performance with LPC pig slurry. Another key positive point was the high methane yield obtained from LPC slurry from pig weaners compared with the reported literature on methane production from pig slurry. Biomethane production from pig slurry/manure, especially from weaners, tends to have a lower methane yield compared with that obtained from this study. The studies by Browne et al. [61] and Miroshnichenko et al. [60] for pig slurries from weaners yielded considerably lower biomethane potentials of 38.0 L/Kg VSfed and 75.5 L/Kg VSfed, respectively [1,60]. On the other hand, studies from Santos et al. [59] and Rodríguez et al. [33] indicated high biomethane yield for pig slurries [33,59]. These studies mostly digested pig slurries from pig fatteners and pregnant sows which tend to yield high biomethane production. The differing biomethane yields for pig slurry/manure seem to be highly dependent on the type of pig (i.e., weaners, fatteners, pregnant sows, and suckling sows) excreting the slurry with another key factor being the type of meal fed to the varying range of pigs. Irrespective, the biomethane yield from the LPC pig weaner slurry performed considerably better than the reported literature on pig weaner slurries and was about 20% less than the highest reported study of various pig slurries/manures.

**Figure 5.** Daily and cumulative biogas and biomethane production profiles for (**A**) soya bean meal pig slurry; (**B**) LPC treatment meal pig slurry; and (**C**) biogas and biomethane potential of treatment and control pig slurry.


**Table 7.** Summary data for biogas and biomethane production from soya bean meal pig slurry and LPC treatment meal pig slurry.

*3.3. Bioenergy Assessment of LPC Pig Slurry and Soya Bean Meal Pig Slurry Feed AD Systems: A Case Study*

To demonstrate the potential impact of the above findings, in this case study, two AD scenarios are considered to ascertain bioenergy production from the usage of LPC pig slurry and soya bean meal pig slurry as feedstock to a community AD plant close to the pig farm where the LPC slurry was produced. This bioenergy assessment consists of mass and energy balance with the major assumptions listed in Table 8 [62]. The results of the bioenergy assessment are displayed in Figure 6 and discussed in Section 4.

**Table 8.** Assumptions made for bioenergy assessment of LPC pig slurry AD and soya bean meal pig slurry AD systems.


**Figure 6.** A bioenergy case study for an AD plant fed with LPC pig slurry and soya bean meal pig slurry.

#### **4. Discussion**

The results of these experiments demonstrate some potential for extended green biorefinery value chains to have positive benefits for the pig sector by supplying LPC as a sustainable alternative to soya bean meal and enhancing the potential for biogas production from pig slurry residues. Pig farmers, like most sectors of agriculture, are under pressure to become more sustainable. It is estimated that 68% of greenhouse gas emissions in the pork-production chain occur at the pre-farm gate phase [63]. A recent study from McAuliffe et al. [64] found that the primary aims of environmental performance improvements in the pig sector are reducing the crude protein content of pig feed and producing bioenergy through anaerobic digestion of pig slurry [64].

From a sustainability perspective, the potential to displace soya bean meal with an indigenous source of protein, such as LPC, could bring some key benefits. Soya bean meal is a major global commodity crop and is widely used in animal feed production, being one of the primary crops cultivated by farming communities across the world. Its production is mainly in the USA, Argentina, Brazil, China, and India, with only a small amount of cultivation taking place in Europe [17]. The soya bean market is closely associated with land use change for agricultural expansion and forest loss, particularly in South America [65]. It is estimated that approximately 2.31 million hectares of forest disappear annually to make way for soya bean production [66]. Despite a moratorium introduced in 2008 to avoid the purchase of soya beans from deforested land, by 2020 a further 133,000 ha of soy in the Amazon, planted on land deforested after this date, has been produced, which is linked to 69 million tonnes of CO2 emissions [67]. A recent study from Franchi et al. [68], applying consequential life cycle assessment (LCA) to compare LPC and SBM for use in poultry diets, found a significantly lower environmental footprint in the case of LPC. A separate consequential LCA study from Parajuli et al. [69] found that a livestock production system that partially displaced Brazilian soya bean imports through integrated green biorefineries coupled with biogas facilities to produce feed and biomethane, and combined crop and livestock production, generated a lower environmental impact compared with the preexisting "conventional" system (livestock production without biogas production or green biorefinery) [69].

Despite the negative impacts, South America still accounts for more than 50% of global soya bean production, of which 70% is exported, with around 21% of these exports coming to the EU for use in animal feed [65]. A recent study from Escobar et al. (2020), mapping carbon emissions embodied in Brazil's soy exports, found that the EU has a significantly higher footprint associated with soya bean imports compared with China; this mainly due to the source of soya beans being linked to deforested areas [70]. To highlight the impact that source can have on potential sustainability impact, referencing the life cycle inventory database Agri-Footprint 6, an economic allocation, and a point of substitution on the system, the footprints for Argentinean-, Brazilian-, and U.S.-sourced soya bean meal are 4.13 kg CO2/kg, 4.28 kg CO2/kg, and 0.53 kg CO2/kg, respectively.

Despite these sustainability issues, the EU is still very dependent on imported soya beans. Soya bean meal is one of the most widely used individual protein sources in Europe, accounting for 29% of crude protein for animal feed in the EU (including the United Kingdom) during the period 2019–2020 [71]. This results in the average European person consuming 61 kg of soy indirectly every year. However, Europe has very low domestic soya bean production, accounting for just 3% of the total demand in the period 2019–2020, demonstrating a low self-sufficiency rate [71]. Overall, the EU's lack of indigenous protein results in approximately 17 million tonnes of proteins being imported on an annual basis, with the majority (13 million tonnes) deriving from soya bean [72]. This has prompted the European Commission to support the development of native protein ingredients such as peas, lupins, and faba beans [72]. On the other hand, grassland is a very available resource across the continent of Europe, with permanent grasslands making up 35% of total arable land use in EU-28 [73]. A study from Mandl [74] investigated the potential of grasslands to deliver additional protein for Europe. Focusing only on 15% surplus grassland, and assuming a yield of 8 t DM/ha, it is estimated that there is a grass surplus in Europe of approximately 20 million t DM/ha, creating an additional crude protein equivalent of approximately 3 million t DM [74]. This potential protein availability can be further increased to unlock more protein from all EU grasslands providing for ruminant and monogastric needs. The integration of legumes alongside grass for biorefining can further increase the sustainability of green biorefinery systems. If we consider, based on the more recent work of this paper, that by biorefining we can give almost 45% of the protein originally present in grass to pigs without reducing the impact on milk production from feeding the press-cake co-product to cows; then, assuming a grass yield of 10 t DM/ha/yr with 20% protein content, we can estimate an LPC biorefinery co-product yield from dairy farming of about 0.5 tonne protein/ha. This would mean that the 3 M tonnes can be obtained from only 6 Mha, or that the 17 M tonnes of protein that we import in Europe can be obtained from 34 Mha, being about half of all grassland in the EU.

Further environmental and self-sufficiency gains may be achieved using this extended green biorefinery model by increasing the potential of renewable energy from pig slurry, with the findings from this study indicating a 26% increase in biomethane yield from slurry produced from pigs on the treatment diet.

In Figure 6, a case study assesses two AD plant scenarios which considers the conventional soya bean meal pig slurry and the current study LPC pig slurry as AD feedstock to a community digester. The bioenergy case study includes the anaerobic digester, a centrifuge for digestate separation and an amine scrubber for biogas upgrading. The community digester can process a high throughput of 48,500 tonnes per annum of slurry, hence this feedstock capacity was assumed for both the LPC and soya bean meal pig slurries. Employing the biomethane yield from (Table 7), 357 and 220 thousand cubic meters of methane is produced per annum from LPC and soya bean meal slurries, respectively. This translates to 3572 MWh and 2198 MWh per annum of renewable bioenergy for the LPC and soya bean meal slurries, respectively. The gross bioenergy from the LPC pig slurry feed AD scenario was about 63% higher than that of the conventional soya bean meal pig slurry feed AD

scenario. Furthermore, considering the parasitic energy demand of both scenarios, only the LPC pig slurry feed AD plant generated a positive net bioenergy of 1010 MWh per annum. This indicated the superiority and efficacy of the LPC pig slurry in utilization to produce renewable energy as compared with the conventional soymeal pig slurry.

According to the European Biogas Association [75], up to 41 billion cubic meters (bcm) of biomethane could be produced from sustainable feedstocks in Europe by 2030, rising to 151 bcm in 2050 [75]. These targets would enable the fulfillment of the European Commission target of 35 bcm biomethane production per annum by 2030, as outlined in the REPowerEU plan. Out of this total, the majority, 38 bcm by 2030 and 91 bcm by 2050, are estimated for anaerobic digestion, with 33% of the contribution anticipated to come from animal manure feedstocks [75].

While the overall CH4 potential of pig slurry as a biogas feedstock can be comparatively low by comparison with other substrates, the utilization of pig slurry for on-farm energy production can have positive impacts on pig farm operations while contributing to a circular economy, also enabling the recirculation of nutrients to local farms [76]. Using LCA to investigate strategies for addressing greenhouse gas (GHG) emissions and fossil energy consumption in European pig production, Nguyen et al. [63] found that using pig manure as a feedstock source for anaerobic digestion had the ability to displace 53% of fossil energy usage and reduce GHG emissions of the farm by 27% over the baseline pig farm scenario [63]. Various other studies have investigated the feasibility and benefits of pig slurry anaerobic digestion for on-farm usage. Freitas et al. [77] investigated the potential of co-digestion with elephant grass silage, or corn silage, as well as the use of a biochar additive in comparison with monodigestion of pig slurry [77]. Using co-substrates and additives enabled greater production of electricity compared with monodigestion, but also resulted in significant environmental impacts associated with co-substrate use, primarily as a result of fossil fuel use during the silage production chain. When comparing digestion with the direct spreading of slurry, Zhang et al. [32] found that digestion of pig slurry alone resulted in a 48% decrease in direct emissions of GHGs (190 tonne CO2e) compared with direct land application, due to the recovery of methane [34]. Using LCA, Jiang et al. [78] compared the co-digestion of pig manure and food waste with alternative management strategies including pig manure land use, food waste mono-digestion, and composting, and found that the co-digestion approach performed better in most environmental impact categories [78]. Using an average size piggery for comparison, comprising 762 sows with 16,000 t/yr pig manure, the same study found that the global warming potential (GWP) only became negative when the inclusion of food waste in feedstock was greater than 2000 tonnes per annum [78]. A further recent benefit of interlinking the green biorefinery approach with pig systems was reported by Regueiro et al. [79] who demonstrated that both unfermented and fermented whey or brown juice from the grass biorefinery can be used to stabilize pig slurry in storage by contributing to reducing the pH. A reduction below pH6 reduces ammonia emissions as well as methane emissions from the stored slurry.

The development of such anaerobic digestion models may also help to meet the needs of the local communities or districts in which they are based. Community-based anaerobic digestion models, which can help to meet the heating and electricity requirements of the community, for example, heating community buildings, are well established. In addition, such models may help to meet the mobility needs of the community, through the development of biomethane-based transport systems in pressurized or liquefied forms. In Europe, biomethane-based transport is at the furthest stage of development in Sweden, where half of the biogas production is used for transport [80], including public transport buses where it supplies the fuel for over 20% of the distance travelled [81]. Biomethane may also help to reduce the emissions and dependency of local manufacturing industries which are heavily dependent on natural gas, including pulp and paper, and some food processing industries.

In addition to the environmental sustainability and self-sufficiency benefits, the development of such local bioeconomy value chains may also help to relieve cost pressures for

local pig farmers. According to the European Commission [82], EU pigmeat production is due to decrease by 5% in 2022 in large part due to rising input costs [82]. This situation has been compounded by the Russia–Ukraine war, as both Russia and Ukraine were key exporters of fertilizers and grain and oil crops, including wheat, maize, and sunflower, and Europe has also been dependent on natural gas from Russia. Between Q2 of 2021 and Q2 of 2022, following the Russian invasion of Ukraine, the average price of goods and services used within agriculture in the EU jumped by 36% for the same agricultural inputs [83]. An economic analysis of small-scale green biorefineries from Cong and Termansen [84] found that using LPC can be economically feasible for both pig farmers and the green biorefinery [84]. In order for these models to be successful and implementable, economic feasibility is an important outcome. The study found that LPC will decrease the average feed cost by 5%. Coupled with this, rising natural gas prices, and a need to reduce dependency on imported Russian gas, has changed the landscape for biomethane in Europe, making biogas more cost-competitive and even cheaper than natural gas in the current environment [85]. This combined with a significant increase in biomethane potential, demonstrated by this study, may offer a greater opportunity for pig farmers to produce, use, and supply renewable energy. While this paper primarily underlines the synergies and benefits that may be found by connecting ruminant and swine sectors in extended local green biorefinery models, as noted earlier, several additional synergies may be found by connecting additional value chains in local bioeconomy models, further improving the resilience of local regions.

#### **5. Conclusions**

Overall, this work demonstrates significant potential for the development of local bioeconomy value chains based on a green biorefinery model, increasing the resource efficiency of grasses and legumes, building synergies, and meeting input requirements to increase the sustainability and resilience of the pig sector. The LPC co-product from the green biorefinery was produced with a high-crude-protein content of 43.9% and has been integrated within the pig treatment diet replacing 50% of the soya bean meal present in the control diet, achieving a slightly higher daily feed intake and average weight gain compared with the control batch on conventional weaner diets. In addition, the slurry produced by pigs on the treatment diet achieved higher biogas and biomethane production rates compared with the slurry from pigs on the control diet. Since sustainability improvements in the pig diet and anaerobic digestion of slurries are identified as key components to increasing the sustainability of pig production systems, a local bioeconomy model may help to meet this objective, while providing additional co-products for use in the ruminant and industrial sectors. Given the abundance of green biomass that exists in Europe, the potential to unlock additional protein from these through local green biorefineries may offer significant potential to increase feed resilience for the pig and broader livestock sector.

**Author Contributions:** Conceptualization, J.G., C.O., D.M., K.D., L.G., S.K., E.B., K.O. and J.P.M.S.; methodology, J.G., C.O., D.M., D.W., K.D., L.G., S.K., E.B., K.O. and J.P.M.S.; formal analysis, J.G., C.O., D.M., D.W., K.D., L.G., S.K., E.B., K.O. and J.P.M.S.; investigation, J.G., C.O., D.M., D.W., K.D., L.G., S.K., E.B., K.O. and J.P.M.S.; resources, J.G., C.O., D.M., D.W., K.D., L.G., S.K., E.B., K.O. and J.P.M.S.; data curation, J.G., C.O., D.M., D.W., K.D., L.G., S.K., E.B., K.O. and J.P.M.S.; writing—original draft preparation, J.G.; writing—review and editing, J.G., T.O., C.O., D.M., D.W., L.A.V., K.D., L.G., S.K., E.B., K.O. and J.P.M.S.; supervision, J.G; project administration, J.G., E.B. and K.O.; funding acquisition, J.G., E.B. and K.O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded through the Farm Zero C project by Science Foundation Ireland's Zero Emissions Challenge, grant number 19/FIP/ZE/7558. The authors are grateful to acknowledge this funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

**Arman Derakhti 1, Ernesto D. R. Santibanez Gonzalez 2,\* and Abbas Mardani <sup>3</sup>**


**Abstract:** In recent years, the Industry 4.0 concept has gained considerable attention from professionals, researchers and decision makers. For its part, the COVID-19 pandemic has highlighted the importance of managing the agri-food supply chain to ensure the food that the population needs. Industry 4.0 and its extensions can address the needs of the agri-food supply chain by bringing new features such as security, transparency and traceability in line with sustainable development goals. This study aims to systematically analyze the literature to address the challenges and barriers against the application of industry 4.0 and its related technologies in the management of an agri-food supply chain. Currently, despite the large number of publications, there is no clear agreement on what Industry 4.0 is, and even less its extensions. The next revolution that includes new technologies and improves several existing technologies brings additional conceptual and practical complexity. Consequently, in this work we first determine the main components of I 4.0 and their extensions by studying the literature, and then, in the second step, define the agri-food supply chain on which I 4.0 technologies are applied. Two well-known databases—Web of Science and Scopus—were chosen to extract data for the systematic review of the literature. For the final evaluation, we identified 24 of 100 reviewed publications. The results provide an exhaustive analysis of the different I 4.0 technologies and their extensions that are applied in regards to the agri-food supply chain. In addition, we find 15 challenges that are classified into five major themes in the agri-food supply chain: technical, operational, financial, social and infrastructure. The four most important challenges identified are technological architecture, security and privacy, big data management and IoT (internet)-based infrastructure. Only a few articles addressed sustainability, which reaffirms and demonstrates a considerable gap in terms of the sustainable agri-food supply chain, with waste management being the one that has attracted the most attention. This review provides a roadmap for academics and practitioners alike, showing the gaps and facilitating the identification of I 4.0 technologies that can help address the challenges facing the efficient management of an agri-food supply chain.

**Keywords:** industry 4.0; agri-food supply chain; sustainability; agri-food 4.0 supply chain; agri-food 4.0; supply chain 4.0; food waste management; water management; agriculture 4.0

#### **1. Introduction**

In recent years, the concept of Industry 4.0 (I 4.0) has attracted the attention of practitioners, researchers and decision makers, and its applications have been studied in multiple industrial sectors. Despite the large number of academic and non-academic publications, there is still no clear definition of I 4.0. Several technologies, such as Radio Frequency Identification, Internet of Things, Cloud Computing are considered I 4.0 components, and in some cases, by themselves define I 4.0. In order to standardize language and set the context for this article, we first tackle the problem of defining I 4.0.

**Citation:** Derakhti, A.; Santibanez Gonzalez, E.D.R.; Mardani, A. Industry 4.0 and Beyond: A Review of the Literature on the Challenges and Barriers Facing the Agri-Food Supply Chain. *Sustainability* **2023**, *15*, 5078. https://doi.org/10.3390/ su15065078

Academic Editor: Claudio Sassanelli

Received: 2 January 2023 Revised: 3 February 2023 Accepted: 20 February 2023 Published: 13 March 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

In 2011 the German government representative used industry 4.0 term as a steadily growing industry that considerably affects our lives. This speech was about how digitalization and new technologies revolutionize the organization of global value chains [1]. As a pioneer country in manufacturing, Germany introduced the idea of integrated industry by launching I 4.0 initiatives in 2011 for its high-tech strategies [2]. I 4.0 is known with different terms in scientific publications such as "Fourth Industrial Revolution", "smart manufacturing", "Industrial internet" or "integrated industry" [3]. Moreover, the other terms suggested by [4] are "smart factories", smart industry", "digital manufacturing", and "smart production". I 4.0 itself is a concept that is applicable through different technologies. The number of technologies is growing and with emerging new technologies it would become more and more. We first determine its main component by studying the literature [3,4] and is more relevant to the supply chain research area.

#### *1.1. Industry 4.0 Key Technologies*

Four items were found as principal components, including cyber-physical system (CPS), internet of things (IoT), smart factory, and internet of services (IoS). The study [4] proposed the four mentioned technologies to introduce a coherent definition of I 4.0. Furthermore, [3] study suggested these four technologies as the fundamental technology components of I 4.0 in logistics which are explained as follows:

1.1.1. Cyber-Physical Systems (CP), and Their Application in the Agri-Food Industry

CPS has been invented to respond to the necessity of developing a connection between the physical and virtual worlds [5]. Thus, CPS is the transformative technology that manages the interaction between computational capabilities and physical assets [6]. CPS accomplishes its goal by using different sensors, communication devices, and actuators. The application of CPS in agri-food has recently been studied in two topics of smart agriculture and smart farming [7,8]. Precision agriculture is one of the achievements of CPS application in the agriculture industry, with more efficient performance and resourcesaving outcomes [9,10]. Precision agriculture is possibly defined as Wireless Underground Sensor Networks, by implementing communication between computers and physical assets with sensors under soil [11] or underground sensor networks to control the quality of soils [11–13].

Furthermore, some papers discuss how CPS application can bring traceability to agri-food systems [14,15]. The application of such systems, in reality, is a challenging task [16–18].

#### 1.1.2. Internet of Things (IoT), and Its Application in the Agri-Food Industry

IoT is considered the main initiator of I 4.0, which became popular in the early 21st century [19]. Physical devices such as equipment, machines, products, etc., are connected virtually at different and remote locations. These items that perform as physical access points are controlled and monitored by cyber systems [20,21]. "Things" are the entities with physical features in a physical property. These "Things" are incorporated flawlessly in a virtual network system that makes IoT, an information system [22].

IoT in the agri-food supply chain helps suppliers and consumers locate products quickly and display product details, leading to choosing fresher products with the help of sensors. IoT helps retailers monitor the food quality and let them waste management of the products that their expire date is close and reduce energy consumption by managing the temperature at the store, freezers, etc. IoT increases traceability, a prerequisite feature for accomplishing previous acts [23].

1.1.3. Internet of Services (IoS), and Its Application in the Agri-Food Industry

IoS might play a key role in the future of industry. Concepts such as software as a service (SaaS), service-oriented architecture (SOA), or business process outsourcing (BPO) are closely associated with the IoS. Barros and Oberle [24] (p. 6) propose a broader definition of the term service, namely "a commercial transaction where one party grants temporary access to the resources of another party in order to perform a prescribed function and a related benefit. Resources may be human workforce and skills, technical systems, information, consumables, land and others". IoS has not been discussed in the agri-food supply chain so far which shows a potential gap for this technology in this area.

1.1.4. Smart Factory and Its Application in the Agri-Food Industry

We have presented CPS, IoT, and IoS so far, which are the main components of I 4.0. The interaction of COS over the IoT and IoS enables a smart factory [3]. Smart factory works in decentralized manufacturing in which "human beings, machines, and resources communicate with each other as naturally as in a social network" [19] (p. 19). Smart factory in the agri-food 4.0 supply chain could be defined as "smart farming" or "smart agriculture".

Smart farming empowers farmers to apply more dependable control. Real-time, on-site processing data reduces time-consuming. Data is transmitted by cloud system for further analysis. IoT devices such as multiple sensors help cover more areas in the remote and expansive coverage areas [25,26]. Data stored within the cloud is also used by processing plants to resolve operational management problems [27].

This study aims to study the I 4.0 challenges in the agri-food supply chain. Therefore, it is necessary to define the agri-food supply chain precisely. Scholars in agricultural economics and management disciplines first proposed agri-food supply chain [28,29]. Agri-food supply chain management, first proposed by a group of Dutch researchers, manages the supply of raw materials for agricultural production, production processing, and product distribution and logistics [30,31]. This term has been mostly used in two research fields agricultural-related disciplines (e.g., agricultural science and agricultural economics), and business management disciplines (e.g., supply chain management and operational research). Based on [32], we consider agri-food supply chain as one of the four terms "agricultural supply chain", "agricultural value chain", "food supply chain", and "food value chain".

The application of I 4.0 in the agri-food supply chain is also known as the agrifood 4.0 supply chain. Furthermore, agri-food sector could be considered in bioeconomy definition [33] which defines an economy based on renewable biological resources.

#### **2. Materials and Methods**

In this study, we applied a systematic literature review (SLR) which was proposed by [34] and developed by [35]. This SLR mainly comprises five steps: research questions definition, search strategy design, study selection, quality assessment, and data extraction.

In the first step, we devise some research questions that should be addressed through this SLR. The questions are regarding the objective of this research. Afterward, in the second step, based on research questions, we come up with a search strategy to find the most relevant publications to the research questions. This step contains both sub-categories: finding the search keywords and determining the literature databases. In the third step, study selection criteria are formed to determine the narrow the most relevant study with respect to addressing the research questions. In the next step, we apply a quality assessment in which we set up some quality checklists to speed up the assessment process. The final data is gathered to answer the research questions in the last step, data extraction.

#### *2.1. Research Questions*

This SLR aims to recognize challenges ahead of industry 4.0 application in the agrifood supply chain. Towards this aim, six following research questions have been formed by the authors:

• RQ1: What classifications of agri-food products have been discussed with the emergence of industry 4.0? RQ1 aims to identify the agri-food products that have used the industry 4.0 context. By answering this question, scholars have a better understanding

of potential research in the agri-food industry, and it demonstrates which products have adapted industry 4.0 technologies compared to others.


I 4.0, with disruptive technologies and interconnected machinery, aims to improve production efficiency, which helps suppliers serve better to their customers. The proposed questions attempt to find out the obstacle against I 4.0 as well as how these technologies address sustainability within agri-food supply chain. In this regard, the first question classifies the agri-food products to find which agri-food products have taken advantage of I 4.0 technologies so far. The second question attempts to find all technologies used in agrifood supply chain and shows which technology has been used in what types of agri-food products to demonstrate the research in this context. The third and fourth questions address the sustainability and aim to find in which manner I 4.0 influences on sustainability of agri-food supply chain. Finally, the last two questions address the challenges and barriers of I 4.0 in the agri-food supply chain and picture different themes which will help practitioners and scholars to have a general picture of challenges from different perspectives.

#### *2.2. Search Strategy*

The search strategy includes three sub-classifications: search keywords and literature databases, which are explained in detail as follows:

#### 2.2.1. Search Keywords

The following steps were done to find the search keywords [35]:


We added the last step as a new important step to address the scope of the research. In the scope of this SLR, industry 4.0 is a widely used concept that has several distinct

technologies that contribute to I 4.0. As mentioned in the Introduction Section, it is necessary to define I 4.0 precisely; otherwise, the research cannot address the research questions properly because there would be many out-of-scope references in the result. To tackle this problem based on [3,4], I 4.0 has four main components, including CPS, IOT, IOS, and smart factory. However, after reading the articles by the authors, we concluded that the terms "smart farming" and "smart agriculture" are used interchangeably in addition to smart factory. Consequently, we added these two terms to the search keywords step. In the search keywords, in order to search for any group of characters before or after a term we add an asterisk (\*). This function is available in both WOS and Scopus.

The search keywords result are shown as follows:

(\*agri\* OR "\*food\*" OR "\*agro") AND ("internet of things" OR iot OR cps OR "cyber physical system\*" OR ios OR "internet of services" OR "smart factory" OR "smart farm\*" OR "smart agriculture ") AND ("supply chain\*" OR " logistic\*") AND ("industry 4\*" OR "I 4.0" OR "agri-food 4" OR "agriculture 4" OR "the fourth industrial revolution" OR "smart manufacturing" OR "industrial internet" OR "integrated industry").

#### 2.2.2. Literature Databases

Two well-known literature databases comprising the Web of Science (WOS) and SCOPUS were used for this SLR. Clearly, the two mentioned search engines are the most famous and completed search engines. This research aims to answer research questions considering the application of the fourth industrial revolution in the agri-food supply chain. Thus, the search keywords were used to search for peer-reviewed publications. The search keywords covered article title, abstract, and keywords provided by the author and publication for both WOS and SCOPUS. As the agri-food 4.0 supply chain concept and its applications are novel, we have not applied any time restrictions in our research.

#### *2.3. Study Selection*

Search phase 1 resulted in 101 peer-reviewed publications (see Figure 1). In both WOS and SCOPUS there is an option that lets the authors filter articles based on their document type, which is a part of the exclusion criteria. This SLR only considers journal articles, conference papers, and book chapters. After applying exclusion criteria and based on the document type criterion, 67 publications remained. We derived all references through the next step and eliminated duplicated publications by Mendeley software. Twelve duplicated articles were deleted through this step. These 55 remaining peer-reviewed publications went through the inclusion criteria, and only 24 articles remained in the pre-final step. In the last step, we applied a quality assessment. Although all publications received different scores, any publications have not been removed through this step and passed this assessment. The authors found out that the number the publications were few and this could be an explanation that is why all of them remained. The other description is that we define the scope of the research accurately, which leads us to a few numbers of articles. In other words, both preciseness of the scope of this research and the novelty of the topic concluded in a few publications but entirely related ones.

The defined inclusion and exclusion criteria are as follows: Inclusion criteria:


Exclusion criteria:

• Duplicates are eliminated by Mendeley and a final revision by the authors;


**Figure 1.** Search and selection process.

As we mentioned earlier, this SLR has two main scopes to find the industry 4.0 challenges in the agri-food supply chain. The first scope is about the I 4.0 that should discuss at least one of the four main components, IOT, CPS, IOS, and smart factory [3]. The second scope of this SLR is the agri-food supply chain, which we defined in the introduction. The included publications in this research should discuss at least one of the four supply chain as follow: "agricultural supply chain", "agricultural value chain", "food supply chain", and "food value chain". Finally, there is no time restriction regarding the novelty of the scope of this research. All publications have been published since 2015 based on our search keywords.

#### *2.4. Study Quality Assessment*

The quality assessment is a process in which we weigh the retrieved quantitative data in meta-analysis [35]. Since the results are too few, a meta-analysis is unsuitable for this SLR. Instead, we only use the result of quality assessment. After applying quality assessment, there would not be any changes in the number of outcomes. There is a reason to justify why there is no reduction in the selected articles after using quality assessment. The number of publications is too few, which shows how accurate and in detail the scope of this SLR is defined.

We devised five main questions and 17 criteria to assess the quality of selected articles which are shown in Table 1. Some of the questions were derived from [36]. Question 1 to question 4 are quantitative that has two answers: "Yes" or "No" which answers are scored as follow: "Yes = 1", and "No = 0". The other 17 criteria are scored in the same manner with answers "Yes = 1", and "No = 0". Question 5 is qualitative which is scored as follow: Excellent quality = 1, good quality = 0.67, Fair quality = 0.33, Poor quality = 0. To calculate the final score for a publication, each question has a weight, which is multiplied by the

determined score for that question or criteria, and finally, we sum the scores. The weights are defined by authors based on their relevance to the scope of this research which are classified into two sub-categories; five main questions and 17 criteria. As the questions have higher importance, they have bigger weights than the criteria. These five questions aim to assess how close the selected articles are to the agri-food 4.0 supply chain area based on the above definition of the agri-food 4.0 supply chain.


**Table 1.** Quality assessment questions and criteria.

Moreover, 17 criteria assess how much the selected publications' general quality is close to scientific research, from article title to references. If a paper obtains 50% of the total score or more, it will be included in the SLR. In this study, all 24 articles received the minimum requirement. The criteria and the final score for each article are shown in Appendix A Table A1.

#### *2.5. Data Extraction*

After obtaining the target articles for answering the research questions, the articles were gathered into a separate excel sheet to find the challenges against the agri-food 4.0 supply chain.

#### **3. Results and Discussion**

This section includes four subsections that present our findings through this SLR. First, we explain the 24 selected articles in terms of publication year, and document type. Afterward, in the following parts, we answer the research questions. The last sub-section contains the answer to RQ5 and RQ6.

#### *3.1. Overview of Selected Articles*

We found 24 peer-reviewed publications that meet our SLR requirements. The first study was published in 2016. The distribution of the publications based on publication year is shown in Figure 2. There is enormous progress in the number of Publications between 2020 and 2021, which shows this area has gained more popularity since 2019. Thus, this Table affirms the novelty and importance of this new research area.

**Figure 2.** Distribution of publications based on year.

Regarding document types, Figure 3. displays the number of Publications according to their document type. The blue color shows the number of Articles which is equal to 18, and conference papers are displayed by orange color, which is equal to 5, and there is only one book chapter among this SLR.

**Figure 3.** The number of articles based on the document type.

The quality assessment result is shown in Table 2. The quality score was calculated in scale 1 and all studies obtained more than 50% and were selected for final SLR. This Table demonstrates that more than 66% of studies have a high or very-high quality level.


**Table 2.** Quality assessment score result.

The distribution of articles in journals is broad, and only Computers in Industry journal has two published articles among all selected studies.

#### *3.2. Types of Agri-Food Products (RQ1)*

This SLR aims to classify agri-food products. In the research [37], a general agri-food product classification was proposed by the authors, including (1) bulk cereals, (2) root vegetables and tubers, (3) sugar and sweeteners, (4) meat, (5) dairy products, (6) fruit, (7) vegetable oils, (8) other. We classified the 24 selected articles into five different categories in terms of agri-food products considering the classification in [37] which are shown in Figure 4. These five classifications include (1) fresh fruits or vegetables, (2) cold chain, (3) packaging, (4) food, and (5) agriculture. In [37], fruits and vegetables are two distinct classifications, however, we consider them in one category since the literature addresses their freshness as well as considering both of them in a single article. Thus, we consider fresh fruits and vegetables one class. Seven Publications refer to fresh fruits or vegetables that seem the most important or the most discussed subject in this area, and it is due to the perishability of these products. The second classification, cold chain, includes fisheries industry meats, or any other material that ship by cold supply chain and refrigerators, in which there are five articles. Packaging as a third classification is a new class we add to [37] because the only two articles that refer to packaging and discuss the importance of packaging in the agri-food supply chain in a general perspective. Finally, we proposed two general categories. Ten remained articles discuss food and agriculture products, respectively. These two terms, "food" and "agriculture" are general terms that do not refer to a specific product. We selected these titles because these two categories contain research papers more associated with general frameworks and review papers. Although these publications discuss the agri-food supply chain they do not concentrate on any specific products. This sub-section and Figure 4 answer the first research question.

**Figure 4.** Classification of publications based on agri-food products based on the number of studies.

#### *3.3. Types of Technologies (RQ2)*

The answer to research question 2 is shown in Figure 5, which demonstrates the distribution of industrial 4.0 technologies based on agri-food products. As we defined before, industry 4.0 has four main Keys, including IoT CPS, IOS, and smart factory.

**Figure 5.** Industry 4.0 technology distribution based on agri-food products.

The authors read all articles through the selected publications and found out that some other technologies play roles in the agri-food 4.0 supply chain. IoT has attracted more attention compared to other technologies. Other technologies that the authors found are big data, robotics, and automation, cloud computing, artificial intelligence (AI), 3D printing, blockchain, augmented reality (AR), cybersecurity, simulation, and virtual reality (VR). Some articles addressed more than one technology in their research; for this reason, the number of technologies is more than the total number of selected papers. Our findings in this section cover nine technologies that were used to systemically review circular supply chains in and application of I 4.0 in the circular economy [38].

#### *3.4. Sustainability Area (RQ3)*

This section presents our findings in terms of sustainability. Only nine articles from 24 selected articles addressed sustainability which shows a gap regarding sustainability in this field. The distribution of sustainability is illustrated in Figure 6. Of these nine articles, only one addressed the social aspect, and all of them addressed the environmental aspect of sustainability.

#### *3.5. The Contribution of I 4.0 in a Sustainable Agri-Food Supply Chain (RQ4)*

This section aims to address research question 4; how I 4.0 contributes to the sustainable agri-food supply chain. When it comes to the environmental aspect of the agri-food supply chain, waste management attracts more attention due to the perishability of products in this value chain. Then, water management has the second step. Finally, only one paper addressed energy consumption management. The information is shown in Figure 7.

**Figure 7.** Areas of sustainability.

#### *3.6. Challenges and Themes in the Agri-Food 4.0 Supply Chain (RQ5 and RQ6)*

This study aims to find challenges in the agri-food 4.0 supply chain. To find the challenges, first, we read all the articles and found four papers that addressed challenges directly for different technologies but not specifically the industry 4.0 concept [13,16]. Then we reread the publications to see which article addresses which challenge. Among the four papers that addressed challenges directly, three [13–16] addressed challenges regarding IoT adoption in the agri-food supply chain. In [39], the authors studied challenges regarding different technologies, including IoT, Robotics and autonomous systems, AI, big data analytics, and blockchain. There are both common challenges and unique challenges mentioned in these four articles. We reread the challenges and tried to integrate the challenges that were mentioned with different names while they had the same characteristics. Finally, we reached 15 challenges and five themes. The themes and challenges are illustrated in Table 3.


**Table 3.** Challenges and themes in the agri-food 4.0 supply chain.

As the number of challenges in this area is extensive, we aim to classify challenges to make them more sensible for scholars and future research. In [55], the article's authors classified IoT adoption challenges into five main themes, including technical, financial, operational, social, educational, and governmental, and fourteen challenges. We kept the four technical, financial, operational, and social themes, and we added a new classification called infrastructural themes. This study considers the educational theme as a subsection of the social theme because the terms education and society have been widely used close to each other. Moreover, the related governmental challenges are a sub-category of an infrastructural problem. We explain each of them separately below.

#### 3.6.1. Technical Theme

After reading, analyzing, and interpreting the selected article, we concluded seven challenges that we consider technical challenges. Each challenge is described separately as below:

• Security and Privacy

Security and privacy have been widely proposed as a significant ongoing issue regarding I 4.0 technologies [43,44], more specifically, security issues associated with IoT technologies [20,23,40]. In industry 4.0, a huge amount of data is generated and distributed through a wireless sensor. Four extensive areas, including data authentication, resistance to attacks, client privacy, and access control regarding security and privacy, must be addressed [20,23]. Network and some parts of this data are about users and their private information, interaction, etc.

In other papers, security is considered cybersecurity [41,45,46] which refers to the security related to cyberspace and the link among physical assets and objects. According to the assertation in [41], insufficient cybersecurity awareness and a lack of educated and skilled people in the cybersecurity area can diminish security and raise risks. Due to an enormous number of stakeholders in the agri-food supply chain, such as farmers, manufacturers, wholesalers, retailers, and consumers, the supply chain is vulnerable to cyberattacks [46].

Blockchain, a fast-growing technology, is proposed as an infrastructure to share data in the agri-food 4.0 supply chain [39,42]. By combining IoT and blockchain, Grecuccio et al. [42] proposed a framework to reduce security in the food supply chain and increase traceability. Although blockchain, like other I 4.0 technologies, is vulnerable to cyber-attack, three key characteristics, including data storage, decentralized network, and peer-to-peer communication, intensify security and privacy considerably [39].

• Wireless power transfer and ambient energy harvesting

This challenge is only presented in [39], which seems a significant issue for applying I 4.0 appliances in the agri-food supply chain. Liu et al. in [39] propose a challenge for recharging small sensors necessary in farms such as underground, underwater, livestock, and trees. Replacing the batteries in these appliances seems unachievable. Wireless power transfer is a good solution by recharging the batteries through electromagnetic waves. However, underground, underwater, and long-distance wireless networks are issues ahead of this solution. Likewise, ambient energy harvesting is another potential solution for this challenge. Some research papers studied and showed that harvesting energy from rivers, movement of vehicles and fluid flow, and the ground surface is possible [39]. Nevertheless, as the converted electrical energy is limited, power conversion efficiency should be further improved.

• Big data management:

Big data has been at the center of attention in the SCM and many studies have addressed it recently. For example, in [59], the authors asserted that big data-driven SCM could facilitate the barriers. Moreover, big data management extensively used with I 4.0 in several contexts. In agri-food supply chain 4.0, it has been mentioned as big data, data management, and big data analytics [47–49]. The devices, appliances, and sensors generate numerous big data that need to be collected, stored, processed, and distributed through different networks. This abundance of data should be managed [20]. Some of the challenges in big data management are data storage, searching, sharing, analyzing, and data visualization. Since the amount of data rises daily, the mentioned challenges need high-end hardware and software and continuous upgrade of the system. The other challenge with respect to big data management is data insecurity [46].

• Reliability, availability, and robustness:

Reliability and availability of IoT devices and services are potential issues in the agrifood supply chain. There might be various failures in the system, such as hardware failure, software issues, malicious attacks, and limited energy. Consequently, the robustness and reliability of I 4.0 services are a significant challenge [20]. Both device failure and link failure, which lead to communication failure, can cause financial risk. In research by [20], authors propose a three-layer architecture for IoT-driven agriculture in which reliability is one of its main features.

• Developing IoT-based cloud system:

A cloud system is used in the I 4.0 context as an intermediary space to link devices and appliances through its network. Yadav and Garg, in their article [40], consider cloud-based development as a challenge ahead of IoT technologies. However, another research [40] considers cloud-based systems pricey and proposes a blockchain system and direct communication of IoT devices through blockchain.

• Technological architecture:

Khan and Altayar [20] propose technological architecture as an obstacle ahead of IoT adoption. They assert that it is necessary to develop, maintain and integrate a robust technological architecture that comprises all IoT-related technologies, such as cloud computing, artificial intelligence, blockchain, wireless technologies, machine learning, big data analytics, and data center and server technologies [20]. Thus, an open technological architecture ensures the integration of different technologies, scalability, mobility, interoperability, modularity, and openness in a heterogeneous environment. This study proves that technological architecture is the most important challenge based on our findings, because it

has attained more attention in the literature than the other challenges in which 50 percent of the selected publications addressed this challenge.

A sustainable food supply chain was designed for IoT-enabled e-commerce food enterprises by applying a lateral inventory share policy [56]. Another study by Almadani and Mostafa [51] presents a systematic integration model of a multi-vendor agricultural production system that uses data distribution service middleware to facilitate communication among production systems.

The cocoa bean traditional assessment method is carried out by humans and takes a lot of time. Adhitya et al. [27] used artificial intelligence classifier methods for textural feature extraction to decrease the assessment time process and cocoa bean waste. Reducing food waste is a contribution to a sustainable supply chain. Jagtap et al. [52] designed a framework to diminish food waste as well as energy and water based on IoT-based devices. Other research [49], studied how cyber-physical systems can contribute to sustainable food systems by using machine learning techniques. Likewise, Mondragon et al. [50], proposed a two-layer conceptual approach in the fishery industry by using IoT devices and digitalization. Furthermore, a software framework that lets IoT devices communicate through blockchain was proposed by [42]. In [48], the author designed a module that uses AI-based IoT devices to ripen the fruits when they are shipped by in the container.

Moreover, Sharma et al. [54] proposed a framework with CPS concentration to improve productivity in the agricultural supply chain. In addition, a three-layer architecture proposed by [42] that is energy efficient, cost effective, heterogeneous, secure, and reliable to use IoT in agriculture based on cloud computing. Finally, the paper by [55] designed a model based on I 4.0 technologies to collect food traceability data through the supply chain and transform it to the consumer to improve the quality of products.

#### 3.6.2. Infrastructural Theme

This theme presents three different infrastructural challenges. These challenges are IoT-based infrastructure, lack of governmental regulations, and standardization. Each challenge is discussed as follows:

• IoT-based infrastructure:

Infrastructure for IoT-based devices and services is another challenge in the agri-food 4.0 supply chain. Accessibility to the internet, the primary service of Industry 4.0, is low in the agri-food sector. In [40], the authors studied the lack of internet accessibility in the Indian agriculture supply chain. The availability and collaboration of a diverse range of services such as AI, CPS, IoT, IoS, blockchain, cloud systems, and so on need a high-quality infrastructure in the supply chain [48]. When it comes to infrastructure, implementation issues arise related to hardware installation challenges and changes in the processes [44]. Likewise, installing these infrastructures needs a high cost [45].

A robust wireless network is necessary to apply I 4.0 technologies in the agri-food supply chain. However, it is an issue because of several causes that negatively affect the wireless network. For example, temperature variations, humidity, human presence, and movements of animals lead to signal fluctuations [39]. Hence, a robust wireless network is vital to coping with weather conditions and the agricultural environment.

• Lack of governmental regulations:

The government plays a key role in the agri-food 4.0 supply chain. Poor regulations in IoT applications may create food safety issues and decrease traceability [60]. In [48], Yadav et al. demonstrated that the collaboration of government organizations, NGOS, and food processing organizations could enable I 4.0 in the agri-food supply chain. Laws and regulations should support the development and extension of I 4.0 in supply chain management, especially in the agri-food sector, which has important consequences on social health and poverty [23].

### • Standardization:

Smart agriculture and global communication need standard protocols to prevent ambiguity and facilitate efficient and smooth integration among different vendors and data safety through cloud networks [23]. Standardization helps devices and digital appliances interact efficiently. This challenge becomes more important when it comes to the global supply chain between various continents with different regulations, responsibilities, and standards [44]. The complexity of I 4.0 technologies even makes it more difficult to define a unified standard for the interaction of sensors, software, devices, actuators, and networks with their own predefined protocols [20]. Lack of standardization has always been a challenge for most new technologies [23].

#### 3.6.3. Operational Theme

This theme attempts to address the barriers regarding the operational aspect of I 4.0 technologies in the agri-food supply chain. The main challenges are high energy consumption, scalability, interoperability, proper connection of ASC entities and IoT technology, and IoT congestion and overload issues. The issues are explained as follows:

• High energy consumption:

Blockchain has gained more attention for this challenge. Due to the high energy consumption of blockchain technology, especially by coal and fossil-based energy, it is a thought-provoking barrier [39]. High energy consumption leads to a high cost of energy that makes it a barrier in the agri-food supply chain. To cope with this issue, scholars in [58] proposed adopting renewable energy to reduce costs in the long-term and meet sustainable development goals for reaching a green planet and mitigating climate change.

• Scalability:

Agri-food 4.0, with growing technologies such as IoT, blockchain, etc., includes an incredible size of devices and nodes that need to connect in the future. Thus, scalability has been mentioned as an ongoing challenge that should be addressed in agri-food 4.0. Some papers presented the scalability issue in IoT [23,40]. The middleware approach is proposed by [40] to provide a flexible service with a huge number of devices that can communicate with each other at one position,. The other research by [44] discusses the importance of servers' scalability in IoT. Scalability in blockchain is presented as an issue in [39] since the transaction speed in blockchain networks such as Bitcoin and Ethereum are so low compared to visa transactions.

• Interoperability:

The nature of I 4.0 works with devices that produce data, and they need to communicate with each other through diverse networks. When it comes to the agri-food supply chain, it contains several layers, including thousands of devices and communication. Therefore, I 4.0 technologies should have the capability to communicate and exchange data between devices. This capability is called interoperability, an issue in the agri-food 4.0 supply chain. The information exchange needs an interoperable environment. Interoperability generally has four types; technical, semantic, syntactic, and organizational [20]. This challenge has been studied for two technologies: IoT and blockchain [20,39]. Different sorts of blockchain networks can hardly communicate with each other, and there is a necessity for interoperable communication protocols [39].

• Congestion and overload issues of IoT:

This issue is similar to the operational challenge proposed by [61] to study collaboration in an industrial symbiosis network in the circular economy. The complexity of the supply chain and collaboration among its players was previously introduced as an issue. Given the complexity of the SCM network in which each layer in SCM uses several devices as well as fast-growing technologies and the existence of big data in the network system, facing congestion and overload of IoT devices is an inevitable challenge in the agri-food 4.0

supply chain. The congestion occurs when multiple devices produce information and want to load their data through the network [40]. It is necessary to find a solution for this issue in the future as an adoption barrier.

#### 3.6.4. Financial Theme

The financial theme has only one challenge regarding general financial concerns in terms of novel technologies as follow:

• High implementation and operating costs:

Implementing I 4.0 requires new technologies and always emerging technologies have been pricy. Furthermore, the maintenance cost will be a challenge for organizations [23]. In a study by [41], active packaging is suggested as a new solution for sustainability in the postharvest food supply chain. Although active packaging increases the quality, safety, and shelf life of packaged food, the high-cost implementation is a barrier to adopting this beneficial tech. On the other hand, the other paper proposes that although the implementation of I 4.0 technologies and digitalization is expensive, in the long term, these technologies are beneficial [58].

#### 3.6.5. Social Theme

We name this theme social because it is about social challenges pertaining to humans. The issues show the lack of human skills and educational issues, which is explained as follow:

• Lack of human skills and educational issues:

It is essential to address the issues related to farmers and I 4.0. Farmers should be aware of the benefits of industry 4.0 context to participate in the supply chain. However, there is a lack of human skills in the agri-food supply chain as most data analysts and data scientists are not at agricultural universities or agri-food-related companies [39]. Likewise, universities have not yet prepared industry 4.0 courses or programs to provide sufficient human skills in this field. Furthermore, the lack of skilled labor is apparent, and it is essential to speed up preparing highly skilled experts and laborers in the agri-food supply chain [58]. In research by [48], education and training were found as the fifth most important enabler of the agri-food 4.0 supply chain among 14 enablers, proving the significance of this challenge.

#### **4. Discussion and Conclusions**

This systematic literature review aimed to address challenges regarding industry 4.0 application in the agri-food supply chain. In the first step, we defined industry 4.0 and its main technology components, including IoT, IOS, CPS, and smart factory. Then we described the agri-food supply chain containing four types of supply chain; "agriculture supply chain", "agriculture value chain", "food supply chain", and "food value chain". Afterward, we devised six research questions. We defined the search terms based on our definition of industry 4.0 and agri-food supply chain. In the final step of the systematic review, we applied a quality assessment. The result of SLR was 24 selected Publications. Although a few articles were selected in the final process (24), the quality assessment showed the selected publications are quite well matched to the scope of this study, which could be due to the novelty of the field and the preciseness of the scope of this SLR. We contribute to the literate by applying a systematic literature review into industry 4.0 application in the agri-food supply chain. A systematic review that concentrates on I 4.0 is scarce. Similar publications mainly discuss specific technologies such as IoT, CPS, or blockchain, and not the I 4.0 as a general concept. Hence, the findings in this study are worth reading, and it helps both scholars and practitioners in the agri-food supply chain.

We found 15 challenges regarding the agri-food 4.0 supply chain that each one is defined separately. Then, we classified them into five main themes. The principal findings of this review are summarized as follows:


Therefore, managing an immense of data is recognized as an issue. Furthermore, all efforts towards I 4.0 are useless without proper infrastructure like the internet.

• (RQ6) Finally, after reading the challenges, we classified them into five main classifications: technical, operational, social, infrastructural, and financial.

The findings of this study showed a considerable gap in the sustainability of agrifood 4.0 supply chain. Meanwhile, most of articles addressed I 4.0 application in the agri-food supply chain without addressing sustainability which draws attention and shows a considerable gap for future research in the agri-food 4.0 supply chain. On the other hand, among articles that addressed sustainability, waste management, and water management were found two sub-sections of sustainability that are more discussed and need to be further addressed. The two issues of water and waste management are the most notable challenges in the sustainable agri-food supply chain. More research is needed to study how I 4.0 is able to address water and waste management. Furthermore, I 4.0 is still emerging with the development of disruptive technologies and is in its early decade.

Moreover, based on agri-food product classification in this study we suggest further research on challenges regarding each product's category because first, this study considered challenges overall on agri-food supply chain and second, the character of each category could make define other challenges or more specific sub-challenges.

Furthermore, this study pictures a comprehensive classification of challenge themes which help practitioners and scholars to narrow the concentration on addressing the problem and to have a better understanding of the whole challenges in one frame. The presented themes outline the main barriers that need to be addressed by the agri-food industry in adopting industry 4.0 into their supply chain.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Data Availability Statement:** No new data were created and Table A1 shows the selected references for this SLR.

**Conflicts of Interest:** The authors declare no conflict of interest.


**Table A1.** Quality assessment

 complete result of the selected studies.


*Sustainability* **2023**

, *15*, 5078

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
