Industry 4.0 in food processing: drivers, challenges and outcomes

2.1 Industry 4.0 adoption and implementation

The Industry 4.0 paradigm introduces a new phase of technology-driven innovation for firms (Wee et al., 2016; Culot et al., 2020). Born in the manufacturing context, Industry 4.0 initially identified a group of technological solutions comprising of the internet of Things (IoT), Big Data and analytics (BDA), augmented reality, additive manufacturing, cloud-based systems, simulation, advanced robotics, horizontal and vertical integration, and cybersecurity (Rüßmann et al., 2015; Trivelli et al., 2019). As there is no universally accepted definition of Industry 4.0, other technologies may be considered depending on the sector, such as blockchain and energy management solutions (Culot et al., 2020). At present, some consensus was reached on the core and central role of IoT, which is at the base of interconnection among the different Industry 4.0 technologies, but also represents the common element among smart devices, plants and systems (Culot et al., 2020). Moreover, IoT sets the fundamentals for vertical and horizontal integration, which respectively allow real-time data sharing and analytics throughout different business functions, and extend this possibility across its borders. However, each technology can be implemented in different business functions, leading to distinct outcomes (Chiarvesio and Romanello, 2018).

In the agri-food sector, Trivelli et al. (2019) have identified six clusters of technologies, namely the monitoring, IoT, automation, decision, hardware and laser, underlining the interrelations among them. For instance, the monitoring cluster (e.g. GPS, GIS) is strongly related to the IoT cluster, which refers to the communication of data among machines and devices (e.g. IoT, RFID, sensor networks). Another complementary cluster includes technologies supporting decision making such as artificial intelligence and data mining. In relation to agricultural entrepreneurship, Secinaro et al. (2022) identified macro-topics and themes referring to the value created through new technologies for agri-food entrepreneurs. However, moving from precision agriculture, extant research has scarcely investigated Industry 4.0 applications across other activities of the agri-food supply chain (Lezoche et al., 2020) and extant works are mainly theoretical. For instance, Rana et al. (2021) have reviewed blockchain applications, whereas Hasnan and Yusoff (2018) have speculated on potential Industry 4.0 applications and impacts in the agri-food supply chain. The scarce empirical agri-food research has so far analysed the opportunities deriving from Industry 4.0 in relation to fresh food products processing in Spain (Oltra-Mestre et al., 2020), the potential technological innovations in an Italian supply chain of milk, fruit and vegetable products and cereals (Saetta and Caldarelli, 2020), the technical challenges emerged during the implementation in production of Industry 4.0 technologies in the case of a UK SME producing water crackers and biscuits (Konur et al., 2021) and the barriers and drivers from a supply chain 4.0 perspective by looking at an Australian food and beverage supply chain (Ali and Aboelmaged, 2022). However, there are no studies that specifically address the outcomes of Industry 4.0 integration in the fruit processing to date.

The selection and implementation of Industry 4.0 technologies is not an obstacle-free process (Wee et al., 2016), and there is no single technological solution fitting all firms' needs (Lezoche et al., 2020). Extant studies have highlighted the existence of perceived barriers and drivers in this process (Müller and Kazantsev, 2021). Barriers can relate to the lack of finance, technological awareness, collaboration, infrastructure, regulations, knowledge and skills, particularly during the selection of technologies (Romanello et al., 2021; Manavalan and Jayakrishna, 2019). During the implementation process, change management resistance and skills requirements become particularly relevant for small companies (Ali and Aboelmaged, 2022; Romanello and Chiarvesio, 2021). Also, some technologies require previous time-consuming processes of knowledge codification and document digitisation that require skilled workforce (Lezoche et al., 2020). Instead, drivers are strongly related to the expected outcomes (Ali and Aboelmaged, 2022; Lezoche et al., 2020), such as increased productivity, flexibility, mass customisation, environmental sustainability, reduction of time and cost to market, and quality improvements (Culot et al., 2020). In addition, Industry 4.0 can also influence value creation and capture mechanisms (Wee et al., 2016; Rüßmann et al., 2015) and product and process innovation (Oltra-Mestre et al., 2020).

2.2 Industry 4.0 expected outcomes in the agri-food context

Advanced technologies in production are expected to dramatically improve plants' and factories' performance (e.g. operating costs decrease, improved operational efficiency, increased productivity, lowering of production costs), while contributing to reduce electricity, water and heat consumption (Borowski, 2021), making productions more economically and environmentally sustainable (Borowski, 2021; Jambrak et al., 2021). Second, in the agri-food context, robotics and automation merged with interconnected machines are expected to increase the levels of food safety and hygiene, while simplifying maintenance and reducing human injuries (Saetta and Caldarelli, 2020; Rana et al., 2021; Hasnan and Yusoff, 2018). In the food processing industry, IoT, alert systems and smart industrial robots endowed with cameras and inspection systems can be used to identify risks and abnormalities, improve food quality (Manavalan and Jayakrishna, 2019) and ensure the fulfilment of food safety standards (Hasnan and Yusoff, 2018). BDA could allow predictive maintenance and diagnosis to service machines and plants (Saetta and Caldarelli, 2020), but also help tracking and forecasting the shelf life of the items and estimating the expiry at real time (Manavalan and Jayakrishna, 2019). Moreover, BDA could help firms finding the right balance between the energy spent and the level of products' transformation and necessary costs, in contrast with the established trend that suggests that companies should try to maximise the shelf life of food products at any cost (Augusto, 2020). In addition, sharing and analysing data across the firm boundaries will increase efficiency, sustainability, flexibility, agility and resilience throughout the supply chain, from farmers to the final customers, contributing to reduce the uncertainty and risks related to products (e.g. shelf life, food quality and safety), processes (e.g. supply lead times, production, resources needed), markets (e.g. demand requirements, price competition) and the environment (e.g. weather, diseases, regulations) in the agri-food sector (Lezoche et al., 2020). The flexibility and reaction speed of industrial and logistical systems merged with BDA could improve the interpretation of customers' demands and needs in terms of tastes and information accessibility (Ali and Aboelmaged, 2022; Rana et al., 2021; Oltra-Mestre et al., 2020; Hasnan and Yusoff, 2018). QR codes and radio frequency identification (RFID) systems can be used to track food materials in the food supply chain, to increase efficiency of stock control (Saetta and Caldarelli, 2020) or to allow consumers to access information on the country of origin and the ingredients included in finished products (Hasnan and Yusoff, 2018). In this respect, blockchain is still unexplored (Rana et al., 2021), but could effectively enhance data sharing, security, anonymity, trust and decentralisation at all levels (Lezoche et al., 2020; Saetta and Caldarelli, 2020), and could increase products' traceability across the supply chain (Ali and Aboelmaged, 2022; Hasnan and Yusoff, 2018; Rana et al., 2021). Blockchain could allow consumers tracking the food's origins and movements across the supply chain, assessing the environmental impact, ethical aspects and quality, with positive returns in terms of consumers' trust, brand reputation and reduced food fraud (Saetta and Caldarelli, 2020). Overall, optimisation through Industry 4.0 is expected to lead to more sustainable data-driven supply chains — supply chain 4.0 (Lezoche et al., 2020) by reducing supply-demand misalignments (Ali and Aboelmaged, 2022).

Last but not least, 3D printing is now used to manufacture small, customised batches of complex metal or plastic products at lower costs. Food 3D printing is still in an experimentation phase (Jambrak et al., 2021; Augusto, 2020). For instance, Foodink is a 3D printing restaurant offering a gourmet experience of 3D printed food (http://foodink.io/). Although food 3D printing for industrial purposes is considered futuristic, it might be promising as a green and clean production technology to produce customised products for specific customers' needs (e.g. vegan).

3.1 Research context

Trentofrutta is a food processing company located in Trento with a factory of 38,000 square meters. With 177 employees and a turnover of about 50 million euro, its main activities include: (1) processing fruits into semi-finished products for industry and (2) contract filling of finished products into bottles and pouches.

Initially labelled Cremogem Spa, the firm was established in 1961. Five years later, the firm changed the name and added a new investor, who gradually acquired stakes in the company until he finally became the sole owner in 1971. The company was then acquired in 1988 by a German group, who, in addition to financial capital, endowed the company with a new name, Trentofrutta, and a strategic business plan aimed at extending the fruit processing operations and expanding the business abroad. Nowadays, Trentofrutta belongs to a family group of enterprises, which includes other complementary medium companies focused on specific fruit processing activities.

The firm mainly operates in the business-to-business context with two main business lines. The first and traditional business encompasses the production of fruit and vegetable-based semifinished products for industry. This business line mainly relies on the establishment of long-term relationships with its customers, such as leading players in the international food industry. The second business line is more consumer oriented, as the firm produces finished products in glass and pouch formats as a white brand for leading beverage and baby food brands and large-scale Italian and European distribution clients. With this second line of business, the firm sells products mainly in Germany, Italy, UK and Holland.

Trentofrutta relies on fruit purchasing, while the rest of the chain is short and internalised (from fruit processing to pouch filling and packaging). The short chain sourcing reflects the decision to ensure high levels of safety and quality and makes the business more environmentally and economically sustainable. Fruit is sourced nationally (except from tropical fruits). The firm has established sustainable fruit supply chains, particularly for biological product lines, while it has a “controlled” chain where firm specialists and agronomists perform strict controls on the use of pesticides among suppliers' and agriculturists' activities. This is fundamental to ensure that semi-finished processed fruit can be used for baby food purposes. Over the years, the firm's policy relied on selecting suppliers and establishing long-term relationships with reputed suppliers (e.g. cooperative firms with long-lasting traditions).

The firm stands out for investments made in digitalisation and Industry 4.0 over the last five years, also in consideration to its medium size. Some implementation processes are still in progress, such as the introduction of additional digital measurement systems in production, new internal data sharing systems, predictive maintenance tools and other innovative technical solutions. Table 1 describes the evolution of investments in digital transformation over the last years.

Since 2015, the firm experienced a change in its vision strategy and the owner and the CEO initiated the digital transformation and firm repositioning in the marketplace. First, they fostered and spread a strong customer orientation, e.g. by developing a new packaging design to respond to customers' requirements. Second, they initiated a process to reposition the firm in a medium-high market segment by introducing the pouch production lines, which were considered more profitable and appealing for customers. Since then, the digital strategy was planned and developed for group enterprises, by leveraging synergies both during the selection and implementation of technologies. As illustrated in Table 2, the digitalisation process started in 2015 with the new management software and information and communication technology systems, followed by advanced robotics and IoT in one plant, a brand-new Industry 4.0 plant in 2017 and big data collection and analytics implemented over the last two years.

Group firms have implemented the same management software and ICT system, which allowed the interconnection of all business functions within the organisation. Production supervision is supported by laptops and screens through which simply checklist activities and uncompleted tasks. Employees use a key card to access departments and record activity developments, which leads to an automatic generation of production and human resource activity reports. Currently, the firm is introducing the use of iPads in production. Through these investments, vertical integration within the company has been successfully achieved. Vertical integration merged with Industry 4.0 in production allow big data collection and analytics at different organisational levels, from orders to production planning and outbound logistics.

A second group of Industry 4.0 investments pertains to production. In addition to existing production lines for glass filling, in 2015 the top management implemented a continuous and automatised pouch-filling production line (not digitalised). As this choice was rewarding, further investments were made to automatise the palletisation phase and widen the tank after some years. In 2017, exploiting the experience gained through the first implementation, the firm introduced also a second production line for pouch filling, completely automatised, robotised and digitalised in accordance with the Industry 4.0 paradigm. This second production line allows automatic controls also related to quality, hygiene conditions and food safety requirements. In general, the firm also adopts X-ray machines, machines to read QR codes, bar code readers, precision weighting machines and other innovative analytical tools in each production line. 3D printing is used to produce spare highly customised parts or single components for cost and saving purpose.

At the moment, horizontal integration with suppliers (e.g. agriculturists) and customers has not been particularly developed yet. In fact, the level of customers' digitalisation tends to be advanced, whereas suppliers are less prepared for this approach. The firm already digitally stores data on the country of origin of products, but customers and final consumers cannot digitally access information. At present, product, process and firm certifications in addition to regular client audits (through audit, inspections, data requests, complete traceability system) are used to guarantee quality to customers.

3.2 Research design

We adopted a case study methodology (Yin, 2009) to explore relevant factors and outcomes related to the Industry 4.0 adoption and implementation processes, which is a relatively new topic in the agri-food sector (Hasnan and Yusoff, 2018; Oltre-Mestre et al., 2020). We chose the case study approach because it is useful to “understand the dynamics present within single settings” (Eisenhardt, 1989, p. 534) and is particularly suitable for theory building purposes, also given the relatively new nature of the phenomenon (Eisenhardt, 1989; Eisenhardt et al., 2016; Lyons, 2005; Ferraris et al., 2019). The fruit supply chain has been already adopted as a context of analysis in Italy (Canavari et al., 2010), but not yet in relation to this research question. Single case studies have the advantage of providing rich descriptions of the object of study and revealing aspects of a phenomenon that has so far been inaccessible (Eisenhardt, 1989; Franceschelli et al., 2018; Yin, 2009). We followed a theoretical sampling to choose the single case study. Through an initial screening of agri-food companies located in Trentino Alto Adige, we identified and purposefully selected Trentofrutta, a fruit processing company located in Trento (Italy), because it has an ongoing digital strategy and an ongoing plan of Industry 4.0 adoption since 2015. This company represented a single information-rich case, which we considered unique and revelatory (Eisenhardt, 1989; Eisenhardt and Graebner, 2007), and, given the ongoing digital strategy, guaranteed the initial assessment of outcomes. It provided an extensive description of what is happening in this particular context of analysis (Welch et al., 2011), reconnecting and enriching previous evidence on the fruit supply chain in Italy (Canavari et al., 2010).

3.3 Data collection and analysis

We chose to collect primary data through in-depth interviews with key managers of the company due to the exploratory nature of our study (Eisenhardt, 1989). We gave voice to informants, who were treated as knowledgeable actors. After consulting a general literature on the topic, we developed an interview schedule based on open questions on the Industry 4.0 adoption to allow the raise of novel insights and facilitate theory building (Goulding, 2002). As Industry 4.0, we included technologies proposed by Rüßmann et al. (2015) and blockchain following agri-food studies (Lezoche et al., 2020; Rana et al., 2021). The interview protocol consisted of five steps. The first in-depth interview was conducted with the Integrated Management System Manager in May 2021. It lasted about 90 min, was recorded and literally transcribed into text. As our objective was to preserve flexibility to adjust interview protocol based on informants responses (Goulding, 2002), iterative interviews were then collected with the respondent to solve and verify unclear issues and integrate data collection. The second interview was collected in July, the third in August and the fourth in September 2021. The fifth step consisted of collecting feedbacks and integrations with the CEO, production director, and human resource and financial director in August/September 2021. We stopped collecting data when we reached theoretical saturation, hence seeing that additional interviews brought minimal incremental learning (Eisenhardt, 1989). Each respondent separately and individually read the case analysis and findings to ensure fidelity of interpretation and for validation purposes (Goulding, 2002). To avoid retrospective bias, we used different sources to gather data for triangulation purposes, including archival documents (e.g. balance sheets, company policy declarations), websites and press releases. Interviews transcripts and archival data were organised in a firm dossier. We used qualitative techniques to analyse data (Eisenhardt, 1989; Miles et al., 2020). After removing data unrelated to the object of this study, we coded data based on a priori codes aligned with the research question (Miles et al., 2020). According to Miles et al. (2020), during data analysis, other codes emerge inductively compared to the initial list. Then, we proceeded with data theming by clustering data according to commonalities, and this facilitated the synthesising and emergence of concepts and relationships between variables (Eisenhardt, 1989; Eisenhardt et al., 2016). This analysis led to the creation of tables and figure included in this article.

4.1 Economic and environmental outcomes

At the economic and environmental levels, automation, digitalisation and Industry 4.0 technologies-led to a variety of outcomes including productivity and operational efficiency increases, and waste, energy consumption and CO₂ emission reduction, increasing the economic and environmental sustainability of the firm. For instance, vertical integration allows to optimise production planning, with positive impacts in terms of waste reduction and energy saving.

The photovoltaic system on the roofs allow to meet average internal energetic needs. We have four steam generators inside the thermal power station and a water purification system. We also self-produce what we need (e.g. spare parts) through 3d printing for instance. This is what characterise our company: we exploit internal technologies and employees’ skills to solve problems, also looking at economic and environmental sustainability. All these plants are interconnected and allow data analytics and energy consumption optimisation. (Third interview)

Compared to the more traditional production line, the Industry 4.0 plant for pouch filling enabled better controls of production processes and throughput time and lead time reduction. Compared to the non-digitalised pouch filling, the Industry 4.0 plant has a lower average operating cost. Automation and digitalisation partially reduced workforce. Compared to the first line, this production plant also includes robots and automatic systems performing 100% controls on quality and hygiene conditions. This improved hygiene conditions in the continuous line, while initial positive impacts were noted on food quality and safety.

4.2 Organisational outcomes

The organisational level emerges in relation to vertical integration across business functions and Industry 4.0 in production, which allowed rapid information and communication flows within the firm and across its borders (e.g. with customers). Digitalisation through the new management software increased flexibility and problem-solving capacity within the organisation, even if requiring organisational and cultural changes as a counterbalance. For instance, digitalisation processes required resource- and time-consuming procedures of knowledge codification. Also, some settings of the new management software were initially perceived as cumbersome and required to be customised and tailored on the necessities of the organisation. This was made possible by the department created ad hoc to support the digital transformation of the firm.

A new business function was created to serve digitalisation purposes: an internal engineering department specialised on technological advancements, in addition to the engineering department for internal maintenance of plants. Both departments benefit from the support of the ICT department, which is located in the group company but serves the whole group. These departments and the digital knowledge and skills accumulated allowed the incremental implementation of other technologies (e.g. BDA), while containing the costs and expenses. This team designed and customised the lines according to specific production requirements and was highly receptive to technological advances.

4.3 Human resource management

As regards the human resource management level, the adoption of Industry 4.0 influenced recruitment processes (new employees, new skills, women, young), current job positions and tasks' characteristics (fewer manual activities, less physical effort, more supervising activities, expected less bureaucracy) and training activities aimed at upskilling/reskilling personnel (organisational culture, digital skills, etc.).

The new hiring procedures, particularly in relation to the Industry 4.0 production plant, resulted into a generational turnover and an increase of female workers. Although employees tend to be reluctant to radical and new digital solutions, young workers showed more adaptability and were able to infuse new competences and skills into traditional practices embedded in the organisation. This became an asset for the firm. Also, operational activities in the Industry 4.0 plant required different skills such as concentration, precision and control rather than hard physical work. Consequently, in production, most processes have now a shift supervisor and a vice-shift supervisor, who share a knowledge and have complementary skills to guarantee that plants can operate even when the supervisor is unavailable. Competency tends to be more distributed in the organisation and new figures in supervisory and coordination tasks were employed. The new hiring procedures resulted in a higher number of female workers in supervisory roles.

The Industry 4.0 production line employees several women particularly in supervisory roles and as “forewomen”, whereas the less digitalised business-to-business production line still sees a higher percentage of men. Also, digitalisation and Industry 4.0 led to create new supervisory roles, which were often assumed by females. (Third Interview)

New roles in maintenance activities emerged and new figures were hired: e.g. the supply chain coordinator and the HR manager. Other tasks are now performed entirely automatically, such as the pasteurisation temperature now constantly verified. As regards the quality function, initial controls of employees are now stronger in terms of machinery settings, but less necessary on the continuous line as machines automatically signal anomalies. These phases have seen a reduction in manual work and an increase in supervision and coordination activities. The use of robots and automation in the riskiest processes increased employee safety, while increasing the work pace. Also, digitalisation allows to trace employee activities, decreasing their daily bureaucratic activities and simplifying their everyday tasks.

The aim is to dramatically reduce bureaucracy for employees, who would be extremely satisfied if they do not need to fill out paper documents anymore (e.g. vacation plan, activity log), and most operations’ results would be digitally generated. (First interview)

Training activities encompassed the strong upskilling and reskilling of personnel, in addition to a program of digital-skills training activities for new employees. In the Industry 4.0 plant, several employees were moved from operational to supervisory activities, and specific training activities were necessary to upskill human resources. Since 2015, the top managers accompanied digitalisation with specific training activities and hired an HR manager, who supervises teams, assesses employees' competences and skills in relation to the assigned tasks, and identifies appropriate training paths. This activity was also supported by an external psychologist, who helped employees to win resistance to change.

4.4 Strategic outcomes

The digitalisation of production timesheets permits the monitoring of time necessary for maintenance, washing and core production, with strong strategic returns in terms of strategic decision making.

Each working hour is detailed in terms of order, working activity, and task, giving top management the opportunity to effectively calculate and analyse the cost of each activity and order. Big data collection and analytics has revealed an exceptional instrument to optimise planning and production, and – overall – to support decision making at the corporate level. (Third interview)

The digitalisation of processes enables the generation of statistics that are used to: (1) accurately plan production and resource allocation among the two business lines, (2) to plan maintenance activities and (3) to support strategic decision making at the corporate level. Top management members use big data analytics to better forecast demand trends and align strategic capacity requirements, and, eventually, mitigate risks related to demand decreases and agriculture uncertainty. This approach allows the organisation to rapidly respond to market requests. Further developments are expected from the additional investments made in data elaboration, statistics generation and digital solutions to extract strategic value from data. Figure 2 describes associations among Industry 4.0 technologies and outcomes.