Exploring Factors Influencing Fourth Industrial Revolution Skills Acquisition in The Food Services Industry

This study explores skills acquisition factors in utilizing 4IR technologies within the Malaysian Food Services Industry, specifically in dine-in restaurants. Applying a Sequential Exploratory mixed methods approach, the qualitative phase explores reasons for using 4IR technologies, factors influencing service staff skill acquisition, and the required 4IR skills in a Food Services working environment. Based on these qualitative insights, a survey instrument for Food Service staff is developed, integrating constructs from the Diffusion of Innovation (DOI) Theory and the Technological-Organisational-Environmental (TOE) Framework. DOI is used to analyze factors influencing technology adoption and skills acquisition, while TOE maps these constructs to technological, organizational, and environmental factors similar to Technical and Vocational Education and Training (TVET) curriculum components. Key findings reveal that 4IR technologies address staff shortages, allowing service staff to perform tasks other than taking orders and sending food to customers. Key factors influencing skills acquisition are the technology’s low complexity and compatibility within the working environment. Key skills include operational handling of 4IR technology, communication skills, risk assessment, escalation of technical issues to maintenance personnel and training of technology use to new staff.


Introduction
It has become commonplace in the recent half-decade that emerging technologies or more specifically, Fourth Industrial Revolution (4IR) technologies are increasingly used in various industries such as manufacturing (e.g.automation, remote monitoring), construction (e.g.Building Information Modelling (BIM), 3D Printing), utilities maintenance (e.g.water and electricity remote monitoring) and various other engineering-oriented industries.However, it has also become apparent that 4IR technologies are becoming more popular and being used in service industries such as Banking, Education, Health Services and currently one of the industries that makes these technologies more easily accessible to the public is the food service industry.4IR technologies which are now increasingly used in the food service industry are such as robot waiters and the Internet of Things via the scanning of QR (Quick Response) codes that store information such as menus of a restaurant.4IR technologies are operationally defined in this study as emerging technologies that are used in the current Fourth Industrial Revolution (since the term was coined by the German government in 2011 Hermann et al (2016) and establish a cyber-physical working environment.Widely referred definitions and discussions on 4IR are such as those by Klaus Schwab Schwab (2016) and in reports by the World Economic Forum (WEF) published in 2016 World Economic Forum (WEF) (2016) and subsequent reports regarding future skills.Subsequent research has delved into the effects of the Fourth Industrial Revolution on the labour landscape.Owing to the technological advancements introduced by the Fourth Industrial Revolution which is characterized by cyber-physical systems, the workforce must possess appropriate skills when utilizing these technologies (Spöttl and Windelband, 2021).
This study focuses on the food services industry, particularly in light of its resilience during the Covid-19 pandemic, where digitization and 4IR technologies played a crucial role such as robotics serving food, IoT-enabled contactless payment, big data analytics for customer preferences, and blockchain with cybersecurity in e-commerce, that showcase the adaptability and integration of these technologies in the Food Service setting.The Food Service sector focused in this study is family dine-in restaurants in Malaysia where food is served by service staff and does not include restaurants in hotels.This is because based on input from one of the interview participants, hotels in Malaysia are not using 4IR technologies to maintain certain elements of the 'human touch' when serving customers who dine at hotel restaurants.
The study aims to explore factors that influence the skills required for a worker in a restaurant that uses 4IR technologies and subsequently identify a skills framework that can be referred to when updating the Technical and Vocational Education and Training (TVET) curriculum.This study operationally defines these skills as Fourth Industrial Revolution (4IR) skills.It is anticipated that the skills required would either be according to the job area, 4IR technology, and occupational level based on the TVET curriculum format in Malaysia in the National Occupational Skills Standard (NOSS).NOSS' counterparts in other countries are such as the Training Packages in Australia, Red Seal Occupational Analyses in Canada, and Workforce Skills Qualifications in Singapore that are developed based on competency requirements (Asia-Pacific Economic Cooperation (APEC), 2014).Research by APEC (2014) on the different forms of TVET curriculum among several APEC economies highlighted that the main components that are common among these curriculum documents are that they are industry-specific, are according to occupational or job levels, and include the components of knowledge, skills, attitude, and tools used.
In the qualitative phase of this study, the factors influencing skills acquisition when using 4IR technologies in the Food Service Industry were explored while keeping in mind the main components of TVET curriculum so that the findings could later be mapped.

Literature Review Fourth Industrial Revolution Technologies Usage In The Services Industry
The advent of Fourth Industrial Revolution (4IR) technologies has transformed the landscape of work environments, incorporating advancements like the Internet of Things, Artificial Intelligence, robotics, virtual reality, and cybersecurity.These technologies, both novel and integrated, play a pivotal role in creating a 4IR-enabled working environment.The Academy of Sciences Malaysia categorizes these technologies as Science & Technology Drivers, encompassing 5G/6G, sensor technology, 4D/5D printing, Advanced Materials, Advanced intelligent systems, Cyber-security and encryption, Augmented analytics and data discovery, Blockchain, Neurotechnology, and Bioscience technology.In the Malaysian context, the Economic Planning Unit identifies foundational 4IR technologies, including Blockchain, Internet of Things, Cloud Computing and Data Analytics, Artificial Intelligence, and Advanced Materials and Technologies.In prior literature, these technologies have been referred to as technology enablers Schwab (2016), technology pillars Ministry of International Trade and Industry (MITI) (2016), and science and technology drivers (Academy of Sciences Malaysia (ASM), 2020).

Theories Relevant To Technology Adoption
Qualitative data collection in this study serves as a precursor to developing a survey instrument in the subsequent quantitative phase.The interview questions are designed to extract responses guiding the survey's construction based on theoretical constructs, representing the independent variables.A Systematic Literature Review conducted by the researchers regarding this study identified the Diffusion of Innovation (DOI) theory (Rogers, 1983) and the Technological-Organisational-Environment (TOE) Framework by Tornatzky and Fleischer (1990) as relevant.DOI highlights innovation characteristics influencing adoption, while TOE extends these to include technological, organizational, and environmental aspects.This study integrates DOI and TOE elements to comprehensively examine factors influencing skills acquisition resulting from technology adoption as applied in previous studies (Haneem et al., 2019;Hiran and Henten, 2020;Zheng and Khalid, 2022).The applied integrated DOI-TOE conceptual framework considers DOI's Innovative Characteristics (Trialability, Observability, Low Complexity) within TOE's Technological Context.DOI's Compatibility and Relative Advantage align with TOE's Organisational Context, analysing technology advantage within the organizational structure.DOI's Social System parallels TOE's Environmental context, examining influences from relevant parties.The resulting theoretical framework is depicted in Figure 1.

Sampling and Data Sources
The overall research applies the sequential exploratory mixed methods approach (Creswell, 2014) which applies a combination of qualitative and quantitative approaches to collect and analyse data according to each sequential phase.The integration of the data occurs in the development of the survey instrument based on the findings from the qualitative phase.This article focuses on the qualitative phase's data collection and results.Data collection was carried out via interview sessions with purposive samples of interview participants representing personnel at the Food & Beverage (F&B) outlets who are either at the managerial level or are in charge of managing the use of 4IR technology at the restaurant.Purposive sampling was applied in the selection of the sample participants with the following criteria: (1) Has experience using 4IR technologies in restaurant work settings; (2) Is a Restaurant Senior Staff (Service Area) or Restaurant Managerial staff or restaurant (franchise) HR personnel or 4IR technology maintenance personnel who work closely with restaurants using 4IR technologies.

Research Instrument
A semi-structured interview protocol was developed and applied as the instrument in the qualitative phase of this study.It was developed based on the research questions of the study and the interview questions were crafted to elicit answers from interview participants on the use of 4IR technologies in food services and skills acquired.The process of validating the interview protocol was done by obtaining confirmation from two experts, one was an expert in the food services industry and another expert was an expert in 4IR technologies e.g.robotics.The semi-structured interview protocol comprised the main questions and probing questions that enabled the researcher to explore deeper on newly surfaced topics during the interview.Table 1 shows the list of Research Questions and sections of the interview protocol that corresponded to the Research Questions.Research Question 1 explores the technologies used by companies and the reasons behind their usage, aiming to grasp the use of the technologies based on the restaurant's objectives.Drawing from Amiron's (2020) findings, companies often apply 4IR technologies to optimize processes and reduce human errors.Research Question 2 delves into the skills essential for workers in the F&B services sector to use such technology.Research Question 3 investigates the factors influencing F&B workers in acquiring skills for 4IR technology use, tied to TOE Organizational and Technological factors.Questions 6 and 7 are not mapped to DOI and TOE, as the questions focus on understanding skills required for using 4IR technologies, which is this study's Dependent Variable.

Data Collection
Data collection involved gathering insights from interview participants, where the number of interview participants followed a method to determine the point of information saturation outlined by (Guest et al., 2020).The number of interviews required was determined by considering the base size (minimum number of interviews to review), run length, and new information threshold.The new information threshold represents the proportion of new insights at a given point in data collection, typically set at 5%, but Guest et al (2020) proposed that researchers could choose between two levels of new information thresholds at either 5% new information or 0% which is no new information.Guest et al ( 2020)'s study suggested that with a base size of 4 interviews and a run length of 2 interviews, saturation could be achieved after 6-7 interviews.Previous studies (Hagaman & Wutich, 2017;Guest et al., 2017) highlight that new information is generated earlier and declines in terms of new information after a small number of interviews.
For this study, calculations indicated an 11.9% new information percentage after the base size of 4 interviews and the first run of 2 interviews, surpassing the 5% threshold.It must added that in the original calculation by Guest et al (2020) there was an overlap of new information between 2 sets of runs, as per their explanation, successive runs overlap, and each set of interviews shifts to the right or "forward" in time by one event.For this study, two additional interviews were conducted, and the new information percentage for this additional run was 2.4%, below the 5% threshold.Consequently, the information from the 8 interviews was considered saturated.The calculations are explained in the following steps.
Step 1 Number of themes for base size of 4 interviews = 42 Number of themes for run length of 2 subsequent interviews 5 and 6 = 5 Therefore, the new Information percentage for the base size and first run length of 2 interviews = 5/42 = 11.9%The new information percentage of 11.9% for the base size and first run length of 2 interviews is more than the new information threshold of 5%, therefore an additional 2 interviews were conducted.
Step 2 New unique themes identified in the additional run of interviews 6 and 7 = 2 ( 1 theme from interview 6 and 1 theme from interview 7) New Information percentage for an additional run against base size = 2/42= 4.7% This result quotient of 4.7 % is less than the threshold of 5%, therefore the information from the 7 interviews can be considered to be saturated.However, as a precaution by the researcher of this study to ensure that there was no missing important information by stopping the interviews upon the 7 th interview, another interview was conducted which was interview 8.However, no new themes were identified.
New unique themes identified in the additional interview 8, and previous interview 7 = 1 ( 1 theme from interview 7 and 0 theme from interview 8) New Information percentage for the additional interview = 1/42= 2.4 % The result quotient of the additional interview was 2.4 % which is less than the threshold of 5%, therefore it shows that the additional interview is still under the threshold of 5% showing that the information was saturated and also proving that new information does decrease over time (Guest et al., 2020).
Table 2 illustrates a total of 9 interviews conducted, including an initial interview with a 4IR technology specialist responsible for installing and maintaining robots in multiple restaurants.This specialist's insights served as a benchmark for comparing findings with restaurant workers.Saturation analysis began with Interview Participant 2 (IP2), forming the base size for interviews from IP2 to IP5.Additional data saturation was achieved in subsequent interviews with IP6 to IP9.These eight interview participants were restaurant personnel using 4IR technologies such as robot servers and the Internet of Things (IoT) for QR Code-based online menus.The list of interview participants and their backgrounds is presented briefly in Table 2. where the initial interview with IP1 served as a source for snowball sampling.However, after obtaining information from the first interview participant, the planning of selecting sample interview participants became more apparent.The subsequent interview participants were selected based on the restaurant's use of 4IR technologies and also type of restaurant.
The saturation of interview responses indicates that their experiences using 4IR technology in restaurants were mostly similar.This may be due to the common working environments and common type of technology used in the restaurants which were robots and QR codes.The interviews were mostly conducted in situ which is that the interviews were conducted at the restaurants while the interview participants were working so the interviewer could also briefly observe how the technologies were used by the workers in the restaurant.

Data Analysis
The data analysis was conducted applying both deductive and inductive approaches.An inductive approach was applied to identify patterns and themes in the interview responses then the deductive approach was applied when mapping the themes and sub-themes to the study's DOI and TOE theories' constructs.Thematic analysis was applied when analysing the interview transcriptions.Saldana (2013) defines themes as phrases or sentences that describe a unit of data and what it means.Theme recognition was applied which is based on the premise that a repetition of codes indicates that it is most likely to be considered as a theme (Guest et al., 2012).However, Saldana (2013) stresses that themes do not depend on the frequency of themes, but also their meaning.Two coding cycles were conducted.The first, open coding, identified recurring themes in the initial 4 interviews and in each subsequent run of 2 interviews, revealing new themes or additional information.The second cycle refined codes into 12 main themes and 48 sub-themes, ensuring comprehensive coverage without omitting important text.The first round was inductive, aligning themes roughly with TOE and DOI theories' constructs.The second round identified new themes and compared them against TOE and DOI constructs for grouping.NVIVO facilitated coding, classifying interview transcriptions as Cases for effective querying, interpretation, and sense-making.The themes, sub-themes and related DOI and TOE constructs which were mapped after the analysis of themes are listed in Table 3. Themes 11 and 12 were not mapped to DOI and TOE as they were mostly about the required 4IR skills.The definitions of the codes elaborated above are included in the codebook to guide consistent coding of transcriptions based on standard definitions operationally defined in the study.To analyse the relationship and frequency of coding in the transcriptions, the crosstab matrix query in NVIVO was applied.The findings from these queries are presented and discussed in the following section, Findings and Discussion.

Findings and Discussion
The findings in this section are presented according to the research questions of the study, where the related findings during the interviews are elaborated below.

Research Question 1 (RQ 1): Why is Fourth Industrial Revolution (4IR) technology used in restaurants?
Research by Amiron (2020) highlighted in its findings that one of the main enablers for the use of 4IR technology in an organisation was based on the organisation's goal of using 4IR technology such as to overcome the shortage of workers or to increase work productivity.Therefore in this study, RQ 1 asks this question to understand the motivation behind the use of 4IR technology in F&B.Based on the findings from thematic analysis conducted on the interview transcriptions with the restaurant personnel, the two most frequent answers were that the use of 4IR technology was an alternative to the shortage of workers that was high during the Covid-19 lockdown period, as a new attraction for new customers and to enable service staff to carry out tasks other than serving customers.Extracts from the interview transcriptions are shared below."..it was because we did not have enough workers, but then it assisted marketing."-IP 4 "At first headquarters used the robot due to lack of workers." -IP 3 "they (the staff) do not have to deliver food as much and they can do other work."-IP 2 Other reasons for using the technology were, as an attraction in areas with a lot of F&B outlets, where the use of robots would be unique in areas with high-end crowds.The latter were mostly reasons provided by privately owned restaurants that implemented the use of 4IR technologies such as robots and QR codes earlier than other restaurants at the time.These restaurant' owners were forward thinking of how they could stand out by using robots as an attraction and according to the DOI theory, they can be considered as early implementors of 4IR technology in the Food Services industry.Extracts are included below." Mainly from customer's experience, the robots attract more customers.This is because the robot helps in marketing, we worry that customers will get the food late but customers like it better when robots send the food." -IP 4 " Yes, it was to be different from the other F&B restaurants because at that time there were no others using robots.and as an attraction as the first restaurant in Johor." -IP 5 The usage of online menus accessed via QR Codes is a form of Internet of Things (IoT) and is one of the 4IR technologies.Based on the interview participant's views, it is used by restaurants due to its advantages which are monitoring customers' food ordering preferences, updating of menu online such as when an item is not available, obtaining customers' details such as phone numbers.A QR code is generated for each table and is useful for online payment and ordering of food, or ordering for delivery.

Research Question 2 (RQ 2) : Which skills are required by workers of the Food Services sector?
The results from the analysis of skills required by the Food Services sector are presented in Table 4 below according to Interview Participants (to observe the frequency of skills mentioned) and skills mentioned in the interviews.Table 4 shows that findings from the analysis of interviews highlight that key skills for restaurant staff include operational tasks that are carried out when handling the 4IR technology such as cleaning and charging of robots and updating of online menu items.Skills that enhance the skill set of these workers are risk assessment and effective communication.
When it comes to technology use, especially with robots, staff must assess and manage risks like spills onto robots or customers.Adjusting the robot's speed based on the type of food or drink being carried is essential for successful delivery to designated tables.Monitoring is crucial for detecting malfunctions, and staff need to promptly communicate technical issues to maintenance personnel through direct messaging.This may lead to temporary pauses in robot use until the issues are resolved, typically taking a few days to a week.
In the context of online menus accessed via QR Codes, communication skills play a vital role.Staff must respond promptly and effectively to customer inquiries, ensuring a positive experience during online ordering.Troubleshooting includes addressing issues like internet connectivity and online menu access, tasks that can be handled by authorized staff.The analysis also revealed that interview participants with more technology experience emphasized the importance of these skills.In areas with stable internet connections and techsavvy customers, the need for assistance with QR codes was reduced.

Conclusion
Based on the findings, the adoption of 4IR technologies initially aimed to address the challenges posed by worker shortages during Covid-19 social distancing requirements.Other advantages became apparent with robots handling repetitive tasks such as sending food to customers, and customers directed to online menus through QR codes, where restaurant staff were able to focus on other tasks.Consequently, the adoption of these technologies necessitated the development of skills among workers, extending beyond operational duties to include effective communication with customers, addressing technical issues, and training colleagues in robot and QR code usage.Notably, skills in handling robots and managing online menus, along with understanding network connectivity and electronic gadget use, became crucial for workers to address customer concerns.
Building on these findings from the qualitative phase, the subsequent quantitative phase will involve the development of a survey instrument.This survey will target restaurant workers using 4IR technologies, aiming to delve deeper into the factors influencing skills acquisition and the specific skills required.This comprehensive analysis, considering both occupational levels and job areas, ensures a direct connection between the study's findings and the design of TVET curricula which in Malaysia is the NOSS.This approach avoids skills mismatches by tailoring curriculum elements to specific occupational areas and levels, ensuring that the skills taught align closely with industry demands.The study's findings enable the mapping of 4IR skills to elements of the NOSS in Malaysia.By identifying the main influences on skills acquisition, which are technological and organizational factors, it is apparent that embedding skills required when using 4IR technologies in the curriculum could be done in NOSS sections that correspond to the technological and organizational factors.In terms of Technological factors, the NOSS has a section that lists the skills according to the main work activities and usage of certain technologies which is named, "Related Abilities", therefore the skills required when using robot waiters or online menus accessed via QR codes can be stated in this section.In terms of Organisational factors, this can be incorporated according to occupational levels and job areas that deal with the use of 4IR technologies, where via this study it has been identified that restaurant service staff and restaurant managers.Identifying the implications of rapid technological advancements on the current workforce is crucial for adapting the curriculum and skills training to keep abreast with the evolving needs of the industry.

Figure 1 :
Figure 1: Theoretical Framework showing TOE and DOI factors Figure 1 shows the factors that influence 4IR Technology Adoption leading to 4IR skills acquisition.

Table 1
Semi-structured Interview Questions Vs.Research Questions and Relevant DOI and TOE

Table 1
aligns interview responses with Research Questions, Interview Questions, and factors influencing technology adoption, contributing to skills acquisition according to DOI and TOE.

Table 2
Profile of Interview Participants

Table 3
Overall View of Codes and mapping to DOI/TOE Constructs

Table 4
Skills Required by Food service staff