1. Introduction
As modern hotel technology, robotic applications were first mentioned in a literature review paper in early 1990s [
1]. In the late 1990s, Graf and Weckesser [
2] pointed out the importance of adopting mobile robots in service applications by testing the feasibility of the mobile robot MORTIME in the hotel environment and proved its applicable dynamic working environments (i.e., hotels). From the 2000s to mid-2010s, although some studies started the investigation of the acceptance of hotel robots, the investigation of robotic applications in hospitality and tourism is very limited. An early example of robotic application was televisions in a hotel environment [
3,
4]. Since 2015, more studies started to investigate robotic applications and human–robot interactions in hospitality and tourism [
5,
6]. For example, by collecting data from expert panel and semi-structured interviews, Kuo, Chen, and Tseng [
7] identified 61 items from the supply side to solve issues such as seasonal employment and 53 items from the demand side for hotels to better transit into service hotels.
Recently, an increasing number of studies, because of the COVID-19 pandemic, explored topics such as the trend of automation and artificial intelligence and the intention of using service robots from the perspective of consumers in hospitality and tourism [
8,
9]. For instance, Tuomi, Tussyadiah, and Stienmetz [
10] identified five roles for service robots including “support” and “substitute” roles through observations and interviews from the perspective of suppliers within the hospitality context. By collecting consumption data from more than 87 countries and regions, Ivanov and Webster [
11] examined the appropriateness of robot usage for tourism-related services and found information provision as their most common usage.
Ultimately, previous studies on robotic applications explored more in the hospitality industry than in tourism. Moreover, most studies explored robotic applications in both hospitality and tourism from the perspective of consumers, while fewer studies examined these from the suppliers’ perspectives. Furthermore, previous robotic application-related studies mainly examined service robots, such as check-in and check-out robots in hotels and food delivery robots in hotels and restaurants. Although the number of studies relating to robotic application in hospitality and tourism has recently been increasing since the pandemic, a comprehensive literature review of robotic applications in hospitality and tourism remains lacking to a certain extent. Hence, to fulfill this research gap, this study comprehensively reviewed 86 articles regarding robotic applications in the hospitality and tourism industries.
Theoretically, this study provides a holistic picture of state-of-the-art applications of robotics in hospitality and tourism settings. This aim is achieved by conducting descriptive analysis, exploring research trends, incorporating contributed journals, delving into research regions, and considering various approaches and applied theories to present a general view of this stream of literature. Moreover, content analysis was carried out to analyze the research content of each study in detail from different perspectives. Additionally, potential research directions were identified for academics on the basis of the review findings. Practical implications on robotic design and adoption for robotic developers and suppliers are accordingly provided as well.
2. Methodology
Following the procedures and standards of existing technology in relevant hospitality and tourism review research [
12,
13,
14], this study conducted descriptive and content analysis to gain a thorough understanding of robotic applications in hospitality and tourism. The details of the research process including data collection, data cleansing, and data analysis are shown in
Figure 1 below.
2.1. Data Collection and Data Cleansing
Relevant robotics-related articles in hospitality and tourism were collected in late May 2022 with all researchers participating in the data collection process. A pilot search was individually conducted by each researcher before formal data collection to identify potential keywords and criteria. A consensus was reached for search keywords and criteria after discussion among all researchers. Next, one researcher performed the initial data collection. The researcher read the titles and abstracts of the displayed publications to determine the relevance of the articles. The collected relevant robotic articles were then passed to other researchers to examine the relevancy.
Following previous hospitality and tourism review articles [
12,
13,
14], three of the largest databases (i.e., Scopus, Web of Science, and ScienceDirect) and one popular search engine (i.e., Google scholar) were selected to access the potential articles. A two-round data collection was performed to extensively retrieve relevant articles. The search criteria were as follows: first, in the initial round of data collection, the used search keyword set was (“robot” OR “robotic” OR “robotic application”) AND (“hospitality” OR “tourism” OR “hotel”). The titles, abstracts, and keywords of articles with each combination of the abovementioned keywords were then retrieved. Second, considering English is the most widely spoken language in the world, it was determined to be the article language selection. Third, only full-length empirical journal articles were collected— book reviews, conference papers, viewpoints, and conceptual papers were excluded. Fourth, no time limit was set for selected article publication dates. The second round of data collection was performed to check any potential missing data. The used search keyword set was (“robot” OR “robotic” OR “robotic application”) and “restaurant”. A total of 446 relevant articles were collected. To ensure data quality and consistency, only articles published in hospitality and tourism social sciences citation index (SSCI)-listed journals were used. After filtering the articles, 97 relevant articles were sent to other researchers for final review. Finally, a total of 86 articles were then selected for further analysis.
2.2. Data Analysis
The selected 86 robotics articles were thoroughly reviewed and analyzed at the data analysis stage. Descriptive analysis was performed to analyze the articles’ published years, research regions, research sectors, perspectives, applied methods, and applied theories. Content analysis was conducted to categorize the themes of selected articles and provide a detailed understanding of the existing robotic relevant literature. All researchers participated in data analysis. Specifically, a pilot data analysis of 10 articles was conducted by two researchers together to generate specific themes. After that, two researchers analyzed the results and assigned the themes to the articles independently. Finally, the coded themes were compared, where the calculated inter-rater reliability was found to be at 0.83, reaching almost perfect agreement with Cohen’s Kappa statistic [
15]. Discrepancies were discussed until a consensus was reached between two researchers.
3. Results
3.1. General Overview of Robotic Applications in Hospitality and Tourism
Figure 2 shows the development trend of robotic relevant publications. As elaborated in
Table 1, 86 existing robotics articles were distributed in 17 hospitality and tourism SSCI-listed journals. The first empirical robotic relevant articles in hospitality and tourism research were published by the International Journal of Contemporary Hospitality Management in 2017 [
7]. Initially, there were only a few robotics-related publications. One was published in 2018 and three in 2019. Surprisingly, the number of relevant publications increased dramatically to 26 in 2020, which is also the first year of the COVID-19 outbreak [
16]. Later, the number of publications rose to 35 in 2021. Until May 2022, 20 robotics articles were published online. Therefore, there is a promising trend in the publication number for the end of 2022. Among the existing 86 studies, the International Journal of Contemporary Hospitality Management, International Journal of Hospitality Management, and Sustainability contributed the most, with a total number of 22, 11, and 11 articles, respectively. The contribution details of other journals are listed by year in
Table 1.
The results in
Table 2 illustrated that 70 (81.4%) of pertinent robotic research studies were conducted in the hospitality context. Within this context, more than half of the existing research studies (
n = 42) were carried out in hotels, 25 in restaurants, and three in both hotels and restaurants. In total, 13 articles (15.12%) were conducted in the mixed context (i.e., both hospitality and tourism). However, only three studies (3.49%) focused on the tourism context.
Robotic applications in hospitality and tourism have become a globally popular topic since 2017. Pertinent research was carried out in more than 17 countries and regions (
Figure 3 and
Table 3, respectively). Taiwan was the first region that was studied in 2017 [
7]. Meanwhile, the United Stated (US) (
n = 19), Mainland China (
n = 18), and South Korea (
n = 13) are the top three most researched regions. A total of 17 articles were examined within a transnational context. Moreover, the research region was not applicable to three articles due to their research nature (e.g., a robot logistics system proposed by Lee, Kwag, and Ko [
17]).
The research perspective and corresponding applied approach are elaborated in
Table 4. A total of 64 (74.42%) relevant robotics articles were examined from the demand-side perspective. Among articles from the perspective of consumers, quantitative analysis (
n = 47) was adopted as the main approach, followed by a qualitative approach (
n = 11) and mixed methods (
n = 6). Additionally, a total of 16 (18.6%) articles were investigated from the supply-side perspective. Likewise, the quantitative approach was adopted as the main approach used (
n = 8), followed by the qualitative (
n = 7) and mixed methods approaches (
n = 1). Meanwhile, six (6.98%) articles were conducted from a multi-perspective (i.e., both the supply and demand side) viewpoint where the mixed methods approach (
n = 5) was frequently used. From the overall research approach aspect, a quantitative approach was most frequently adopted by researchers, totaling 55 (63.95%) articles. A total of 19 (22.09%) articles adopted a qualitative method, and 12 (13.95%) articles adopted a mixed methods approach.
Table 5 presents the distribution of each method used in pertinent robotics research. For the quantitative approach, a survey (
n = 33) was the most used method, while 20 articles performed experiments. Three articles adopted a mathematical or econometric technique to perform the data analysis. As for the qualitative approach, a big data analytics on-online review (
n = 9) and in-depth interview (
n = 9) were the ones most frequently adopted. Observation was used for three studies (
n = 3). Both a case study and the Delphi technique were conducted only once (
n = 1). The total frequency of adopted methods was higher than the total number of adopted approaches since several articles adopted more than one quantitative or qualitative method. For example, Wong, Huang, Lin, and Jiao [
18] adopted a qualitative approach by conducting both big data analysis of online reviews and in-depth interviews to explore the background of smart service in restaurants.
On theoretical foundations, 39 pertinent robotics articles were data driven while 47 were theory driven (
Table 6). Among these theory-driven articles, 31 applied only one theory while 16 of them applied two or more. For instance, Zhong, Coca-Stefaniak, Morrison, Yang, and Deng [
19] applied both the technology acceptance model and perceived value theory to investigate consumers’ acceptance of robots before and after the pandemic. More than 40 theories were applied in relevant robotics articles. The most popular applied theory was the technology acceptance model (
n = 12), uncanny valley theory was adopted as the theoretical foundation in four articles (
n = 4), while cognitive appraisal theory was applied in three studies (
n = 3).
3.2. Articles from the Perspective of the Demand Side
3.2.1. Anthropomorphism and the Preference of Consumers
Sixteen articles discussed anthropomorphism and consumers’ preferences on robotic servers where experimental design was the main method used. Among these 16 articles, seven focused on consumers’ preference for different robot features. In terms of appearance, echoing the uncanny valley theory, non-humanoid service robots were more acceptable than humanoid robots among hospitality and tourism consumers [
20,
21,
22]. Particularly, compared to human-like robots, consumers were more willing to be served by moderated humanoid robots such as machine-like or animal-like robots [
23,
24]. However, robots with some human-like features such as voice and language style can lead to more positive service consequences than those with machine-like features [
25]. Seo [
26] found that female humanoid robot staff members were more attractive than their male counterparts.
Four articles noted that anthropomorphism influences consumers’ expectations of the service experience. In the hospitality context, an anthropomorphizing robot server received higher expectation from consumers [
27]. For instance, humanoid robot chefs were perceived to cook better than non-humanoid ones in the restaurant [
28]. Meanwhile, mascot-like service robots were expected to bring positive emotions to the consumers while machine-like robots were more hardworking [
29]. Robots that are perceived as warm were more acceptable in hedonic service settings, while those that appeared competent were expected to serve in utilitarian settings [
30].
Along with the comparison among robots, five articles (
n = 5) compared consumers’ preference on human and robotic service. Although suppliers and designers have tried to make the robots more attractive and to customize their service, most consumers were more satisfied with human employees than robot employees within the context of usual hospitality and tourism [
31,
32,
33,
34]. However, when the pandemic outbreak was most salient, consumers would choose robotic service rather than human service [
35].
3.2.2. Consumers’ Perception of Robots
Other than anthropomorphism, 16 other articles focused on consumers’ perception of robots. However, exploratory approaches such as the qualitative analysis of online reviews and interviews were frequently adopted as well. Among all articles in this theme, 11 (
n = 11) presented consumers’ experience of robotic services in hotels and restaurants. For example, by conducting interviews and surveys, Zhang, Balaji, and Jiang [
36] identified six common robot features perceived by consumers: role, competence, social presence, warmth, autonomy, and value facilitation. Moreover, based on online reviews, Huang, Chen, Huang, Kong, and Li [
37] summarized consumers’ experience with robotics into four dimensions (i.e., sensory, cognitive, affective, and conative). Two other articles discussed different perceptions from different groups. Lee, Lee, and Kim [
38] indicated that the perception of robotics may differ among different groups such as the ordinary, the enthusiastic adopter, tech laggard, and the value seeker. Meanwhile, by performing content analysis of online reviews, Choi, Oh, Choi, and Kim [
39] found that consumers from different culture backgrounds correspondingly valued different aspects of robot services. Specifically, the Japanese value emotional interaction while the non-Japanese seek functional value. The remaining three articles compared consumers’ experiences of different robotic services. For instance, chatbot resulted in more service failures than self-service technology and online ordering systems in hotels and restaurants [
40,
41]. Meanwhile, consumers had a lower expectation of robot-delivered food than human-delivered food [
42].
3.2.3. Influential Factors of Robotic Adoption
Investigation of the antecedents of robotic adoption accounted for the largest area of research from the consumers’ perspectives (n = 25). Here, most articles conducted surveys to test hypotheses established based on relevant theories (e.g., technology acceptance model, stimulus–organism–response theory, and perceived value theory). Influential factors of robotic adoption can be divided into four categories: robotic, human–robot interaction, human, and contextual.
Specifically, robotic influential factors refer to the characteristics of service robots. For example, robot appearance, competence, and performance impact consumers’ robotic adoption intention [
43,
44,
45,
46]. Influential factors related to human–robot interaction refer to consumers’ perception or expectation of the interaction experience. For instance, consumers’ perceived usefulness, perceived ease of use, perceived value, and trust of robot staff positively influence their adoption intention [
47,
48,
49,
50]. Meanwhile, the social presence of robots and good rapport between a human and robot lead to a higher acceptance of robotics in the hospitality and tourism industry [
51]. Human-related influential factors are the features of consumers themselves. These include previous robotic service experience, personal habits, and innovativeness which have effects on consumers’ robotic adoption intention [
52,
53]. Furthermore, service settings play an important role in affecting consumer’s behavioral intention. For instance, due to social withdrawal tendency effects, consumers were more willing to adopt robotic services than human services when in crowded destinations [
54]. Zhong, Coca-Stefaniak, Morrison, Yang, and Deng [
19] found that under the background of health-related crises such as COVID-19, hotel guests’ acceptance of robotic servers greatly increased in comparison to otherwise normal contexts.
3.2.4. Consequences of Robotic Adoption
Seven articles focused on the positive consequences of robotic adoption. The experimental results showed that hotel service robots had a positive impact on customers’ booking intention and experience, especially during the pandemic [
55,
56,
57,
58]. Additionally, based on user-generated data, Çakar and Aykol [
59] and Mariani and Borghi [
60] found that the greater the interaction between customers and robots, the higher the customer’s positive perception of the hotel. Furthermore, social feedback from robotics can remind customers of eco-friendly behavior in hotel rooms [
61].
3.3. Articles from the Perspective of the Supply Side
Although suppliers are the key players in robotic application, there is a lack of attention on this group of stakeholders. In total, 16 articles were conducted from the perspective of suppliers, which can be categorized into three themes: robotic application and management (n = 6), suppliers’ perception of robots (n = 4), and antecedents and consequences of robotic adoption (n = 6).
3.3.1. Robotic Application and Management
Hospitality and tourism marketing and management strategies continuously evolve along with technical innovation. In particular, service practices have become increasingly automated with robotic assistance. However, human–robot cooperation remains a key issue affecting the success or failure of said service delivery. Therefore, six articles explored the partnership between robots and human staff and provided some managerial implications from the suppliers’ perspective. Adopting both observation and in-depth interview, Tuomi, Tussyadiah, and Stienmetz [
10] summarized the roles of service robots in hospitality and tourism sectors into five roles, namely, their support, substitution, differentiation, improvement, and upskilling. Meanwhile, Tuomi, Tussyadiah, and Hanna [
62] identified the extrinsic and intrinsic drivers and the four layers (contextual, social, interaction, and psychological) influencing suppliers to implement robotics. Using the analytic hierarchy process, Jabeen, Al Zaidi, and Al Dhaheri [
9] also indicated five factors influencing automation in the hospitality and tourism industry: human knowledge, services, robotics, and the internal and institutional environment. However, the adoption of robotics can be either positive or negative. For example, the implementation of robotics replaces some jobs but simultaneously creates additional tech-savvy jobs as well. Thus, to better integrate robotics into service encounters, Tuomi, Tussyadiah, Ling, Miller, and Lee [
63] proposed a decent work-through automation model that includes the effectiveness of human–machine cooperation, working conditions, and level of empowerment. Moreover, to optimize the deployment of service robots, the operating algorithm of the robots is continuously optimized by academics [
17,
64].
3.3.2. Suppliers’ Perception of Robots
Using exploratory or inductive methods, four articles investigated how hospitality and tourism suppliers such as hotel managers and employees perceived their robotic partners. Adopting a mixed-method approach, Ivanov, Seyitoğlu, and Markova [
65] performed a survey and an interview among Bulgarian hotel managers on their perception on robots. Although robots can replace human staff to do some menial tasks, managers still believe that robots reduce the overall service quality, with the UK- and Turkey-based suppliers presenting a similar perception [
66,
67]. While hoteliers acknowledge robots’ capacity to greatly improve service efficiency, the adoption of robotics is a risky investment due to uncertain costs, challenges in hotel culture, and human–robot communication issues. Through in-depth interview with Chinese hoteliers, Fu, Zheng, and Wong [
68] identified five reasons why employees’ resist their adoption of robots: inauthentic looks, low usability, excessive workload, technical insecurity, and robotic uncertainty. In summary, from the supplier perspective, there are both pros and cons in adopting robotics in the hospitality and tourism industry.
3.3.3. Antecedents and Consequences of Robotic Adoption
A total of six articles conducted a deductive study to examine the influential factors and consequences of robotic adoption. Parvez, Arasli, Ozturen, Lodhi, and Ongsakul [
69] performed structural equation modeling (SEM) based on the technology acceptance model to examine the effects of perceived usefulness and ease of use on suppliers’ intention to adopt robotics. Furthermore, Pizam et al. [
70] used the technology–organization–environment framework and identified three positive influential factors affecting suppliers’ robotic adoption intention: perceived advantage, competitive pressure, and top management support. Robotic adoption is a risky decision because employees’ robotic awareness is positively related to their turnover intention [
71]. However, support from the management level and positive competitive environment between humans and robots can attenuate employees’ turnover intention [
72]. Meanwhile, based on cognitive appraisal theory, perceived risk and challenge appraisals are, to a certain extent, positively related to employees’ work performance [
73,
74].
3.4. Articles from a Multi-Perspective
A total of six articles were conducted from a multi-perspective viewpoint. First, three articles [
75,
76,
77] examined the influential factors of robotic adoption from the perspective of both consumers and suppliers. Contributing to the discourse on technology acceptance model, Pillai and Sivathanu [
75] collected the interview data from experts of the tourism industry and then tested the influential factors of the perceived ease of use, usefulness, intelligence, and anthropomorphism on the adoption of chatbots. Van et al. [
76] also found that antecedents of a technology acceptance model, a trustworthy system, and value-enhancing service affect consumer willingness to adopt robotics. Moreover, Qiu, Li, Shu, and Bai [
77] indicated that the rapport between customers and human staff positively mediate the relationship among service robots and customers’ hotel experience.
Second, three other articles [
7,
78,
79] discussed various robotics-related topics, including the perception of multi-stakeholders, robotic management, and education. Choi, Choi, Oh, and Kim [
78] conducted two studies among hoteliers and customers. Experimentation was performed based on the themes extracted from the interviews with suppliers. Results indicated that hotel customers perceive human employees to perform better in interactive physical service encounters than their robot counterparts. From both supply and demand perspectives, Kuo, Chen, and Tseng [
7] conducted a SWOT analysis on the development of robotic applications in the Taiwanese hospitality industry. Finally, for sustainable development of robotics in the hospitality and tourism industry, Bilotta et al. [
79] updated the content of tourism technology courses, which cultivates tourism students’ awareness and skills of related techniques.
4. Discussion
According to the review results of 86 articles, a sudden increase in publications on robotic application in hospitality and tourism since the outbreak of COVID-19 (i.e., 2020) is apparent. Three hospitality and tourism SSCI-listed journals, i.e., the International Journal of Contemporary Hospitality Management, International Journal of Hospitality Management, and Sustainability, responded quickly and contributed most to this stream of literature. The US, Mainland China, and South Korea are the most popular relevant robotics research regions. Among all articles, more than 80% of extant research was conducted within the context of indoor hospitality (i.e., restaurants and hotels). Survey and experiments were the main methods adopted by academics. More than 70% of existing relevant articles focused on the consumer perspective but lacked necessary attention on supplier attitudes and responses to robotic application. Despite the undeniable scholarly effort, the knowledge development of robotic applications in the hospitality and tourism is still in its nascent stage. Therefore, existing research gaps and future directions are identified based on these findings.
Publications from the consumer perspective accounted for the largest portion of existing research, with anthropomorphism being a popular topic. Experiments were frequently carried out to test consumers’ preferences for robotic appearance (e.g., humanoid or non-humanoid). However, few studies have investigated consumers’ preferences for other features of robotics. Therefore, it is expected that subsequent studies investigate consumers’ preference for materials, colors, volumes, and voices of service robots. Furthermore, the antecedents and consequences of robotic adoption were extensively studied by SEM. Theories such as the technology acceptance model and perceived value theory have also been used as the basis of several studies. However, existing robotic studies that follow deductive approaches are usually only conducted in specific populations. Hence, additional efforts should be made to compare the differences between populations from various backgrounds. Additionally, extant research focused on the positive consequences of robotic adoption. However, service failures brought by service robotics must not be overlooked. Thus, more attention can be geared toward studying the negative consequences and solutions of robotics service failure.
Knowledge of robotic application from the perspective of suppliers remains relatively limited. Existing studies only provide an initial understanding of suppliers’ perception on robotic application and human–robot collaboration. However, considering suppliers are the key stakeholders in implementing and managing robots, it is crucial to understand their strategies and difficulties while adopting robotics. For example, it remains unknown whether leadership style is indeed an influential factor in managers’ resistance or acceptance of robotic servers. Thus, more statistical evidence from the supply side is expected to test the antecedents and consequences of robotic applications. Furthermore, the way suppliers deal with service failures brought by robotics is an important issue given its continuous popularity. Moreover, rapport building of human–robot teams is a tricky issue that should be discussed more in the future.
There are few robotic articles from the multi-perspective viewpoint. However, to achieve successful and sustainable robotic application, it is important that both suppliers and robots provide services according to the actual needs of customers. Thus, to avoid mismatch between the supply and demand sides, it is necessary to conduct studies that link the opinions of both suppliers and consumers.
5. Conclusions
Although robotic application has recently attracted increasing attention from both academics and practitioners, very few articles present the whole picture of existing robotic knowledge. Therefore, this study aims to provide a comprehensive review of robotic application in hospitality and tourism by conducting the descriptive analysis and content analysis of 86 robotic research studies.
Specifically, the trends of the number of publications, contributed journals, research regions, sectors, adopted approaches, and applied theories were analyzed through descriptive analysis. General information about robotic applications in hospitality and tourism research was also provided, where the number of robotic-relevant publications dramatically increased since the pandemic. Among several hospitality and tourism SSCI-listed journals, the International Journal of Contemporary Hospitality Management contributed most to the existing literature. The US, mainland China, and South Korea were regions which academics conducted the most robotic research. A quantitative approach, including the experimental design and survey, was the most frequently used research method, with the technology acceptance model being the most popular theory adopted. Although suppliers are the key players of robotic application, most publications nonetheless focused on the consumers.
Moreover, articles were categorized by different themes. Content analysis was performed to analyze article details based on different perspectives (e.g., suppliers, consumers, and multi-perspective). Four main sub-themes transpired from the consumer perspective, namely, anthropomorphism and the preference of consumers, consumers’ perception of robots, influential factors, and the consequences of robotic adoption. Meanwhile, three sub-themes from the supplier’s perspective, namely, robotic application and management, suppliers’ perception of robots, and antecedents and consequences of robotic adoption, were also revealed. Research gaps were then identified accordingly.
5.1. Theoretical Implications
This study provided several theoretical contributions to the realm of robotic application in hospitality and tourism. The publication volume of robotics in hospitality and tourism research has increasingly grown in recent years. However, an overall understanding of the current state of knowledge development remains to be developed. Thus, this study conducted a thorough review of extant robotic applications in hospitality and tourism research and provides a profound knowledge foundation for this stream of literature. This study presented the publication trend, contribution of journals, research regions, methods, and applied theories of existing research on robotic applications in hospitality and tourism. Moreover, following the review findings, research gaps and potential research directions are identified from different perspectives, which can provide references for academics in the future studies. For example, studies on consumers’ preference of robot’s features, comparative studies among different cultures, and consequences and solutions of robotics service failures can be extended in the future. Additionally, to optimize the robot service and connect the supply and demand sides, it is advocated that more studies should be conducted from the multi-perspective.
5.2. Practical Implications
In view of these research findings, practical implications of robotic application and management are provided for robotic developers and hospitality and tourism practitioners. Several issues should be considered from the aspects of consumers, employees, and robots.
To better meet the expectations from consumers, suppliers should investigate the actual needs of the demand side before implementing robotics. Surveys and interviews can be conducted to investigate the preferences of hotel guests for robot features and functionality. Preliminary market research helps managers understand market demands and reduce unnecessary investment costs. Simultaneously, in the initial stage of robotic applications, it is recommended to supplement human services with robotic services. For example, human staff members who are well trained in robotic systems can provide demonstrations for consumers in need (e.g., less tech-savvy or elderly people). To gratify the employees, managers should strive to create rapport between human staff and robot staff. Although robot servers are efficient and automated, employees still harbor negative emotions about their robot co-workers. Thus, managers should provide front-line employees with necessary organizational or emotional support to reduce their anxiety about being replaced by robots. Moreover, employees should receive more training in technical skills to collaborate with robotics. At the level of robots, to ensure the efficiency and make good use of robotics, developers should optimize the operational system periodically. Finally, when robotic applications become more common in the future, the ethical issues of robotics should receive more attention.
5.3. Limitations and Future Research
Although this study provides a comprehensive picture of existing robotic applications in hospitality and tourism research, several limitations abound that can be improved in future studies. First, only SSCI-listed journal articles were selected to be analyzed in this study, and academics can consider expanding journal criteria to expand the review results. Second, this study did not investigate the types and work roles of the robots. Therefore, scholars are encouraged to differentiate the robot types for further investigation in the future. Third, this study only analyzed the robotic-relevant publications until May 2022. A periodic review on robotic application in hospitality and tourism is necessary in future studies. Finally, this study advocated more studies on robotic applications in hospitality and tourism. Future studies can build the co-citation network, identify the bibliometric indicators by quartiles to reflect the impact of the cited articles for science, produce word clouds by exploring the similarities, and examine the descending hierarchical classification of articles and the grouping of textual corpus words through systematic literature reviews proposed in previous studies [
80,
81,
82,
83].