Sustainable Logistics: Exploring the Determinants of Consumer Attitudes and Intention to Use Toward Autonomous Delivery Services

Abstract

Autonomous delivery services offer significant cost-saving benefits for logistics and e-commerce companies while helping mitigate environmental issues and urban traffic congestion. As these services become increasingly prevalent in China, understanding how user participation influences adoption decisions is crucial. This study analyzes attitudes and intention to use toward autonomous delivery by integrating factors such as sustainability and customer participation within the technology acceptance model. Notably, prior research has rarely included consumers aged 50 and older, despite their growing engagement with online shopping and delivery services in China. To address this gap, the present study incorporates data from this demographic to enhance representativeness. Employing SEM combined with multi-group analysis through AMOS 24.0, this study examined 526 valid survey responses. The results demonstrate that attitudes toward autonomous delivery systems are positively shaped by perceived usefulness, ease of operation, sustainability features, self-efficacy beliefs, and customer engagement levels. Moreover, user attitudes emerge as a significant predictor of behavioral intention to adopt such services. Additionally, technology-related anxiety and personal innovativeness levels were found to moderate the attitude–intention linkage. These insights establish a theoretical framework to guide future advancements in autonomous delivery system development, marketing strategy optimization, and academic exploration within this domain.

1. Introduction

The exponential expansion of online retail has propelled logistics firms to pursue innovations in last-mile delivery solutions [1], denoting the terminal phase of supply chain operations where goods are transferred to final consumers [2]. This segment commonly represents the costliest, least efficient, and most problematic component within logistics networks [3]. Furthermore, road transportation, which serves as the predominant logistics mode, contributes substantially to worldwide carbon emissions and environmental transformations [4]. To address these challenges, automated delivery solutions are gaining traction. Technologies such as drones and autonomous delivery robots help mitigate ecological concerns, reduce costs, improve accessibility, and ease traffic congestion. These innovations enhance efficiency while addressing issues such as misdelivery and labor shortages, making them a promising solution for future logistics [5,6].

As these technologies mature, their adoption is expanding across cities worldwide [7]. The retail sector, in particular, is expected to benefit significantly from robotics, creating substantial commercial opportunities [8]. Logistics companies and policymakers are increasingly recognizing the cost-saving potential and efficiency improvements of autonomous delivery systems. Consequently, the use of drones and autonomous delivery robots is gradually increasing [4,9]. Robots are playing an increasingly crucial role in transporting goods from warehouses and distribution centers to retail stores and customers. Technological advancements and market developments have rendered this domain exceptionally active, as evidenced by ongoing trial initiatives and operational deployments across European, American, and East Asian territories. Notably, China’s market demonstrates distinctive prominence within this global landscape [10].

China’s autonomous delivery industry is experiencing rapid development and technological advancement. By the end of 2023, over 2000 autonomous delivery vehicles had been deployed nationwide [11]. JD Logistics alone operates over 600 autonomous delivery robots across 30 cities in China, reflecting the continuous integration and expansion of this technology in the commercial sector. Additionally, several domestic manufacturers are actively promoting this trend, including companies like Neolix (Beijing, China), which produces L4-level autonomous driving vehicles, as well as Unity Drive (Shenzhen, China), GoFurther AI (Changsha, China), and White Rhino Zhida (Beijing, China). Among these, White Rhino Zhida has announced plans to produce and deploy 5000 autonomous delivery robots in the market within the next five years [10]. In the drone sector, China’s low-altitude economy surpassed 5 trillion yuan in 2023 and is projected to reach 20 trillion yuan by 2030 [12].

Autonomous delivery robots (ADRs) are robot systems designed to achieve contactless package delivery and are typically sized similarly to humans. ADRs are generally classified into two types: Sidewalk Autonomous Delivery Robots (S-ADRs) and Road Autonomous Delivery Robots (R-ADRs). S-ADRs are primarily used on sidewalks and bicycle lanes. Due to their small size and limited speed (approximately 6–8 km per hour), they are well suited for narrow, densely populated urban environments. Most S-ADRs have a delivery range of up to 20 km and can carry goods weighing between 50 and 100 kg, depending on the model. Their lower speed also results in shorter stopping distances, thus enhancing safety in crowded areas. In contrast, R-ADRs are larger and more robust, designed to operate on roadways and capable of working alongside cars. They generally travel faster and have greater loading capacities, making them suitable for medium- to long-distance deliveries, particularly in suburban or rural areas. Delivery drones (unmanned aerial vehicles) are unmanned aircraft that can operate autonomously or be remotely controlled. In terms of drone technology advancement, China holds the strongest competitive market, followed by the United States [13].

In this study, autonomous delivery services (ADS) are defined as a delivery method utilizing autonomous delivery robots or drones. This includes a range of autonomously performed delivery services, such as food and product delivery.

The procedures for ADS by Alibaba (Hangzhou, China), JD (Beijing, China), and MEITUAN (Beijing, China) are similar. The delivery driver loads the items into the unmanned vehicle at the delivery station in advance and inputs the recipient’s address into the system. The system then automatically plans the optimal delivery route. Upon arrival at the recipient’s location, the package can be retrieved by entering the password from a text message, scanning a QR code, or using facial recognition. Customers can also schedule unmanned vehicle visits through the application to receive deliveries [14].

From the above, it can be deduced that, whether using unmanned delivery vehicles or drone delivery, users must go to a designated location to collect their goods. Given the operational benefits of contactless delivery systems and recent progress in AI, digital infrastructure, and transportation innovations, enterprises are increasingly likely to adopt automated service solutions as substitutes for human labor, strategically positioning themselves for enhanced market competitiveness in evolving commercial landscapes [15]. However, when consumers have limited service options, they may feel that unmanned services are being imposed on them [16]. In particular, in service areas targeting the general public, such as retail, some consumers (e.g., those accompanied by children) may feel compelled to use self-service systems when they actually require human assistance [17]. ADS are expected to become a major delivery method in the future because of their multiple advantages, including addressing ecological concerns, reducing costs, improving accessibility, and alleviating traffic congestion [6]. However, whether consumers are ready to adopt ADS remains a topic worth exploring in depth.

Building upon these considerations, the current research seeks to investigate determinants shaping user perceptions of ADS and their behavioral adoption intentions, while concurrently evaluating regulatory influences exerted by technology-related apprehension and individual innovativeness levels. In pursuit of these objectives, the investigation formulates specific investigative inquiries as follows:

• RQ1. What factors influence consumer attitudes toward ADS?
• RQ2. Does consumer attitude affect the intention to use ADS?
• RQ3. Do technology anxiety and personal innovativeness act as moderators in the association between consumer attitude and intention to use ADS?

By examining these issues, this research establishes a theoretical basis for ADS adoption and strategic diffusion while empirically validating the proposed strategies. Furthermore, the findings not only offer valuable insights to help consumers utilize ADS more effectively but also support the widespread adoption and popularization of these services.

2. Theoretical Background and Hypothesis Development

2.1. Technology Acceptance Model

Emerging from Fishbein and Ajzen’s Theory of Reasoned Action [18], the technology acceptance model (TAM) was formulated to elucidate user adoption intentions toward technological systems and identify behavioral drivers. This framework postulates that behavioral intention, which is determined by usage attitudes, functional benefits (perceived usefulness), and operational simplicity (perceived ease of use), directly governs actual system utilization. These two cognitive constructs exhibit a mutual influence in shaping user attitudes. Specifically, functional benefits denote users’ assessment of technology’s task-enhancing capabilities, whereas operational simplicity denotes the degree to which users consider that system interaction requires negligible exertion [19].

Owing to its extensive applicability within technology adoption studies, the TAM has been extensively employed and further developed through academic explorations. Since its introduction, the model and its extensions have been empirically applied to various new technologies, including mobile technology, gaming, and electric vehicles [15,20]. ADS represent a technological advancement that differs significantly from traditional delivery methods. Studying the behavioral determinants of this new delivery approach within the TAM framework is well justified. Therefore, this study aims to conduct an in-depth analysis of consumer attitudes and intentions regarding ADS by integrating additional relevant variables into the TAM framework.

Despite gaining widespread adoption for explaining and forecasting behavioral patterns and practical system usage, the TAM faces critique regarding its simplistic structural framework [21]. Extrinsic motivational elements driving human behavioral objectives have been conceptually linked to perceived operational utility and interface accessibility [22]. However, Davis’s model has been criticized for overemphasizing extrinsic motivation while overlooking intrinsic motivation [23]. Intrinsic motivation, a key psychological concept associated with voluntary exploration and curiosity, plays a crucial role in cognitive development and has recently gained increasing attention among developmental robotics experts [24].

To overcome the TAM’s theoretical constraints, scholarly efforts have augmented the model through the integration of intrinsic drivers including interaction-based pleasure and individual innovativeness [21,25]. Building upon these developments, the current investigation enhances the TAM’s explanatory capacity regarding consumer perceptions and behavioral inclinations toward ADS through the inclusion of perceived enjoyment as an intrinsic motivator. This psychological construct denotes the inherent satisfaction or hedonic value emerging from user engagement with technological implementations [26].

Within the research framework, perceived usefulness is operationalized as consumers’ assessment of ADS capabilities in optimizing retail logistics efficiency, particularly regarding the temporal expenditure reduction and unpredictability mitigation inherent in traditional delivery methods. Perceived ease of use denotes user evaluations of ADS interface intuitiveness, operational fluency, and absence of specialized training requirements. Perceived enjoyment encompasses the hedonic satisfaction derived from consumer interactions with ADS implementations.

Park et al. [27] and Weijters et al. [28] analyzed the relationships between perceived usefulness, ease of use, enjoyment, and user attitudes by applying the TAM to self-service technology (SST). Their findings showed that usefulness, ease of use, and enjoyment all had a significant positive effect on user attitudes toward SST. Similarly, Lien et al. [29] applied the TAM to SST and found that usefulness and ease of use significantly influenced air passengers’ attitudes toward adopting self-service technology. Guan et al. [30] conducted a study on the acceptance of autonomous vehicles and found that perceived usefulness and ease of use negatively influenced negative attitudes. In another study related to SST, Shim et al. [31] empirically examined attitudes and intentions to use SST using the TAM, confirming that ease of use and enjoyment positively affected attitudes toward SST.

Based on prior studies, the following hypotheses were formulated:

H1. 

Perceived usefulness positively affects attitude toward ADS.

H2. 

Perceived ease of use positively affects attitude toward ADS.

H3. 

Perceived enjoyment positively affects attitude toward ADS.

2.2. Sustainability

Amid escalating global apprehensions regarding accelerated natural resource consumption, persistent wealth disparities, and corporate accountability in social governance, sustainability has evolved into a pivotal subject within managerial academia and organizational operations throughout recent scholarly discourse [32]. The UN’s Agenda for Sustainable Development emphasizes collaboration among governments, businesses, and society to promote global prosperity while protecting the planet. From a corporate perspective, sustainability is not just about responding to organizational goals or stakeholder demands but also involves integrating sustainability principles into strategic decision-making [33]. As companies grow and consumer awareness of sustainability increases, corporate social responsibility plays an increasingly significant role in consumer decision-making [34].

The logistics industry identifies “last-mile delivery” as the costliest, least efficient, and most problematic phase within supply chain operations [3]. Meanwhile, road transportation, serving as the primary logistics mode, constitutes a significant source of worldwide greenhouse emissions and environmental transformations [4]. In response, ADS are emerging as a potential solution to mitigate the escalating climate crisis by offering advantages such as cost reduction, shorter delivery times, and decreased energy consumption [35]. Therefore, this study considers sustainability a key antecedent factor influencing consumer attitudes toward ADS.

Sustainability, acknowledged as a multifaceted construct, integrates economic, social, and environmental dimensions within its operational framework [36]. Within last-mile delivery systems, environmental sustainability centers on carbon footprint mitigation, while social sustainability emphasizes employment generation, health advancement, and life quality enhancement. Concurrently, economic sustainability prioritizes cost reduction, delivery cycle acceleration, energy efficiency optimization, and traffic decongestion [37]. Consequently, this research investigates the role of sustainable practices in shaping consumer perceptions of ADS.

As consumers become increasingly concerned about environmental issues, many researchers emphasize attracting environmentally conscious consumers by highlighting an eco-friendly image rather than the overall image of an organization [38]. Mathew et al. [39] demonstrated in a study on drone-based food delivery services that an eco-friendly image positively influences consumer attitudes toward drone delivery. Similarly, Klein and Popp [40] found that environmental sustainability positively affects consumer attitudes toward last-mile delivery methods. Furthermore, Edrisi and Ganjipour [41] found that environmental concerns have a significant positive effect on consumer attitudes toward using automated pedestrian delivery robots.

Based on prior studies, the following hypothesis was formulated:

H4. 

Sustainability positively affects attitude toward ADS.

2.3. Customer Participation

Customer participation, often referred to as co-production and co-creation [42], describes the psychological state that emerges when customers engage in creating experiences through interactions with companies and other customers. The essence of customer participation lies in value co-creation through interactive experiences [43]. It reflects the extent to which consumers take on active roles in providing and producing products or services [44]. From a consumer behavior perspective, participation involves investing mental or physical effort [45], which demonstrates their willingness to contribute voluntarily [46].

Customers have long played a crucial role as active participants in service delivery. Engaging customers in service production is an effective way to enhance company productivity and reduce costs [47]. In SST environments, positive customer experiences depend on active and effective participation, making them key elements of value co-creation [48].

Today, SST has become a standard in modern markets. Customer participation in SST services is defined as proactive behavior [49], reflecting the effort that customers invest in service co-production. Such participation not only enhances service quality but also strengthens relationships between companies and customers. Particularly in the SST context, customer participation is considered a core step in the value co-creation process, as customers play a highly active role while service providers take a more passive role [50].

Barki and Hartwick [51] identified two dimensions of user participation. One dimension involves user engagement in development activities, while the other pertains to the specific tasks and activities that users perform during the physical design and implementation process. Based on these dimensions, this study defines customer participation as the extent to which consumers perceive their involvement in the provision of ADS. This participation includes consumer support for autonomous delivery technology, as well as feedback and collaboration in interactions with companies. In other words, it reflects consumers’ willingness to provide opinions to service providers and actively engage in the operation of ADS.

When consumers experience a high level of participation in service delivery, they perceive themselves as being integral to the process. They take pride in positive experiences and tend to exhibit greater tolerance when service failures occur. Moreover, they develop expectations regarding the value they contribute through their participation and the significance of that value [49]. These interactions can manifest as actual behaviors or as emotional and cognitive investment, reflecting the value co-creation that consumers experience in their interactions with companies. In SST environments, consumers must actively participate in service provision, which ultimately shapes their overall service experience [47]. In the context of ADS, consumer participation begins with the decision to choose ADS over traditional manual delivery methods. This participation continues throughout the service process, where users receive delivery notifications, specify pickup locations, travel to designated pickup points, and complete procedures such as scanning QR codes [14]. Furthermore, consumer participation may extend beyond the delivery itself, as users may voluntarily provide feedback or suggestions about the service [51]. User participation is a key determinant of user attitude, as those who engage in the process are more likely to view the system as positive, important, and personally relevant [52]. Accordingly, this study examines how customer participation influences the formation and evolution of attitudes toward ADS.

Empirical research across disciplines demonstrates that consumer engagement exerts a substantial favorable influence on attitudinal dispositions and associated behavioral outcomes. Han et al. [45] demonstrated that participants exhibiting a heightened propensity for engagement displayed an increased willingness to exchange perspectives and concepts, resulting in enhanced online word-of-mouth outcomes. Chen and Wang [50] demonstrated in a study on online airline check-in systems that customer participation fosters a positive perception of value co-creation, which, in turn, enhances customer satisfaction. Lee et al. [53] reported that, in SST contexts, customer participation has a positive effect on attitudes. Similarly, Lee et al. [54] found that, in e-commerce services, customer participation has a positive effect on attitudes toward customization experiences. Pattnaik and Shukla [55] conducted a study on the continued usage intention of PBS services and found that user participation positively influences the intention to continue using the service. Yang et al. [56] noted that, in gamification design, participation behavior has a significant positive effect on attitudes toward brands. Furthermore, Yi et al. [49] further revealed that consumer engagement in service production processes within SST-enabled environments significantly enhances user satisfaction levels. Cumulatively, these findings highlight the critical function of participatory consumer roles in attitude formation, satisfaction improvement, and behavioral pattern facilitation across service delivery systems.

Based on prior studies, the following hypothesis was formulated:

H5. 

Customer participation positively affects attitude toward ADS.

2.4. Self-Efficacy

Self-efficacy is a fundamental concept in psychology and serves as a core self-regulation mechanism based on Bandura’s theory of self-efficacy [57]. Self-efficacy denotes an individual’s conviction in attaining target objectives through personal agency. Such psychological confidence serves as a determinant of decision-making patterns, goal orientation, motivational intensity, resilience in challenges, stress adaptation capacity, and susceptibility to stress-related disorders [58]. Moreover, self-efficacy is a key characteristic of consumers and a central concept in social cognitive theory [59]. It is considered a critical success factor in tasks involving computers, information systems, and SST [60].

Consumers with low self-efficacy tend to avoid complex or unfamiliar technologies, whereas those with high self-efficacy feel confident in their ability to complete tasks and persist in achieving them. High self-efficacy can effectively reduce resistance to technology adoption [61]. Within this investigation, self-efficacy is conceptualized as users’ perceived capacity to comprehend, utilize, and adjust to ADS operational protocols. Prospective service adopters evaluate their technological proficiency, with individuals exhibiting elevated self-efficacy demonstrating enhanced exploratory behaviors regarding system functionalities, proactive information acquisition, and the willingness to interact with technological innovations.

Furthermore, self-efficacy plays a crucial role in shaping motivation, attitude, and intention [62]. Hsiao and Tang [63] demonstrated that self-efficacy is a key determinant of SST adoption, significantly influencing user attitudes. Park et al. [61] explored the role of self-efficacy in AI-based SST adoption attitudes and found that it positively affects user attitudes. Hajiheydari and Ashkani [62] investigated mobile application user behavior in developing countries and revealed that self-efficacy plays a vital role in shaping user attitudes and behaviors. Ling et al. [64] incorporated self-efficacy into an extended mobile technology acceptance model and validated its significant positive effect on attitudes toward mobile investment platforms.

Based on prior studies, the following hypothesis was formulated:

H6. 

Self-efficacy positively affects attitude toward ADS.

2.5. Perceived Risk

Originating from Bauer’s [65] seminal 1960 work, perceived risk has been extensively applied within consumer behavior investigations across marketing and psychological studies [66]. When the outcome of a product or service is uncertain, users may experience anxiety, and this uncertainty or lack of trust contributes to perceived risk [67]. Perceived risk constitutes consumers’ apprehension regarding uncertainties encountered during product/service acquisition, combined with anticipated adverse consequences [68].

Perceived risk is a key factor influencing consumer attitudes, with higher perceived risk generally leading to more negative attitudes toward SST [69]. As a result, it is frequently included as a variable in models such as the Unified Theory of Acceptance and Use of Technology (UTAUT), TAM, and Diffusion of Innovations (DOI) [38,70]. Studies on delivery robots indicate that consumers express primarily two concerns: first, accidents that may occur during operation, such as damage to people, objects, or products due to equipment failure or collisions, which could disrupt deliveries; and second, privacy concerns, including the potential leakage of sensitive information such as the delivery destination or personal data, which could expose consumers to malicious actors [38,41,71]. This study aims to investigate the role of perceived risk in shaping consumer attitudes toward ADS, focusing on these two key concerns.

Within this research framework, perceived risk is operationalized as users’ psychological apprehensions and cognitive uncertainties regarding ADS-related operational hazards, including technical malfunctions, vehicular collisions, parcel integrity breaches, and data security vulnerabilities. Empirical evidence from Sung and Jeon [72] indicates that perceived risk exerts substantial influence on consumer perceptions toward automated service systems. Corroborating this, investigations into autonomous delivery mechanisms reveal that service reliability concerns and logistical failure risks exert adverse effects on consumer adoption intentions [41]. Complementary research by Yoo et al. [38] establishes that performance deficiencies, delivery inaccuracies, and privacy compromises collectively diminish consumer acceptance in unmanned aerial vehicle delivery contexts.

On the basis of prior studies, the following hypothesis was formulated:

H7. 

Perceived risk negatively affects attitude toward ADS.

2.6. Attitude and Intention to Use

Attitude refers to an individual’s consistent tendency to respond positively or negatively to a particular situation or object [18]. This tendency influences behavioral intentions and ultimately shapes their actual behavior. Therefore, consumer behavior can be effectively predicted by measuring attitude, intention, and satisfaction [19]. Ajzen’s [73] theoretical framework posits that behavioral attitudes embody an individual’s cognitive appraisal regarding the desirability of executing specific actions. Intentions denote motivational precursors underlying behavioral execution, reflecting both propensity and anticipated effort expenditure toward targeted activities. Empirical evidence confirms a positive correlation between intention strength and behavioral enactment probability [73,74]. From the consumer’s perspective, autonomous delivery services (ADS) are perceived as a more sustainable alternative to traditional delivery methods, as they help reduce carbon emissions and alleviate urban traffic congestion [5,6]. Additionally, consumers are very sensitive to delivery-related costs such as shipping fees [75]. As more logistics companies adopt autonomous delivery technology, competition among low-cost delivery options is expected to intensify, potentially leading to more attractive pricing offers for consumers [76]. The coexistence of ecological advantages and economic efficiency contributes to fostering a positive psychological inclination toward ADS among consumers, thus promoting a more favorable attitude. Within this investigation, attitude is conceptualized as users’ favorable psychological disposition toward ADS adoption, while usage intention captures consumers’ propensity and anticipated adoption probability regarding future ADS engagement.

In the TAM, attitude serves as a crucial mediating variable between belief and behavior [77]. It not only significantly influences behavioral intention but also affects actual behavioral choices [78]. Regarding ADS, consumer attitude plays a key role in determining technology acceptance at both the individual and organizational levels [79]. Therefore, this study explores the factors shaping consumer attitudes toward ADS based on the TAM.

Shim et al. [31] emphasized that consumer attitudes toward SST significantly impact the intention to adopt SST. Eastlick et al. [80] found that, in retail payment self-service technology, consumer attitudes toward co-production services positively influence their intention to use the service. In the field of autonomous delivery, studies by Edrisi and Ganjipour [41] and Yoo et al. [38] showed that positive consumer attitudes toward both pedestrian and drone-based delivery robots increase adoption intentions. Furthermore, Leong and Koay [74] found that, in drone-based food delivery services, consumer attitudes play a crucial role in shaping behavioral decisions, further reinforcing these findings.

On the basis of prior studies, the following hypothesis was formulated:

H8. 

Attitude positively affects the intention to use ADS.

2.7. Technology Anxiety

Technology anxiety can be classified into two types: general technology anxiety and computer anxiety. Computer anxiety refers specifically to anxiety related to using personal computers, whereas general technology anxiety encompasses a user’s overall psychological state toward various technological tools. This includes fear, anxiety, and anticipation when considering or using technology, making it a significant topic in SST research [81].

Technology anxiety is generally regarded as a negative consumer emotion toward SST. However, it does not necessarily mean that individuals will avoid technology entirely. Instead, it represents the likelihood of experiencing fear when encountering new technology, which can lead to excessive caution when operating equipment or forming negative evaluations of related technology. This anxiety may lead to behaviors such as minimizing interaction time with new equipment or even avoiding it altogether [82]. Although SST offers benefits such as flexibility and time saving, technology anxiety can be a major barrier to its adoption and is a key predictor of SST usage behavior. Therefore, its role must not be overlooked when evaluating consumers’ adoption of SST [83]. Given that ADS are a new SST, consumers are likely to encounter certain psychological barriers prior to usage. Accordingly, this study considers technology anxiety a key variable.

Technology anxiety has been extensively documented as a critical moderator affecting the attitude–intention linkage within diverse technological contexts. Jeng et al. [84] demonstrated that technology anxiety significantly moderates the relationship between attitude and behavior in the context of elderly-oriented smart health wearables. In mobile learning environments, Huang et al. [85] identified anxiety related to mobile technologies as exerting notable impacts on learners’ attitude–adoption correlations. Vo et al. [86] investigated consumer responses to mobile augmented reality (MAR) applications, establishing technology anxiety’s moderating role between attitudinal dispositions and usage intentions. Parallel findings by Giao et al. [87] in digital financial services confirm technology anxiety’s pivotal moderating function within attitude–behavioral intention paradigms.

On the basis of prior studies, the following hypothesis was formulated:

H9. 

Technology anxiety moderates the relationship between attitude and intention to use ADS.

2.8. Personal Innovativeness

Personal innovativeness is a key concept in the DOI theory, explaining differences in individuals’ responses to new ideas, methods, or technologies. These differences stem primarily from varying levels of innovativeness among individuals. Personal innovativeness is a continuous trait that reflects an individual’s natural tendency when encountering innovation [88]. As a consumer trait, it highlights that, while some individuals instinctively take risks and explore new technologies, others remain skeptical and prefer maintaining the status quo. This fundamental difference persists regardless of the specific characteristics of a technology and significantly influences adoption decisions [89].

Perceived risk is an inherent part of adopting innovations [90]. Because of differences in personal innovativeness, some individuals perceive greater risks when adopting new technologies or ideas [91]. Highly innovative users are more likely to be early adopters, acting as “opinion leaders” in the early stages of technology diffusion and facilitating the spread of new technologies to a broader audience [92]. Therefore, in this study, personal innovativeness is defined as an individual’s tendency to accept, experiment with, and adopt new technologies or ideas.

Research in behavioral science and psychology identifies personal innovativeness as a crucial factor in technology adoption and decision-making [93]. According to the study by Wu et al. [94], personal innovativeness plays a positive moderating role in the relationship between usage intention and actual usage behavior. Vu et al. [95] analyzed technology assimilation in developing economies, identifying personal innovativeness as a moderator of the attitude–behavioral intention nexus. Their results indicate that individuals with elevated innovative traits demonstrate stronger attitude–intention conversion efficiency, thereby expediting technological proliferation. In e-commerce contexts, Ahmed et al. [96] established that innovativeness amplifies attitude–purchase intention correlations, with technologically progressive consumers displaying a heightened propensity to transform favorable perceptions into transactional behaviors. Parallel investigations by Lee et al. [97] on digital travel platforms in South Korea corroborated innovativeness’s moderating effects on attitude–adoption dynamics. Collectively, these findings underscore innovativeness’s catalytic role in reinforcing attitude–behavior linkages, positioning it as a pivotal determinant in consumer decision architectures.

On the basis of prior studies, the following hypothesis was formulated:

H10. 

Personal innovativeness moderates the relationship between attitude and intention to use ADS.

2.9. Research Model

Figure 1 presents the proposed research model. This study explores how attitudes are shaped by perceived usefulness, perceived ease of use, perceived enjoyment, sustainability, consumer participation, self-efficacy, and perceived risk. It also examines the effect of attitude on intention to use and assesses the moderating roles of personal innovativeness and technology anxiety.

3. Materials and Methods

Data for this study were collected using the WENJUANXING survey platform, combining online surveys with a voluntary sampling method. The research purpose was explained at the beginning of the questionnaire, along with a detailed introduction to ADS and their key features. This study adopted voluntary sampling instead of traditional probability sampling. Voluntary sampling involves selecting respondents who willingly participate and meet the research criteria within the target group [98]. While this method may introduce selection bias and limit generalizability, it is a practical alternative when obtaining random samples is challenging [99]. Thus, voluntary sampling was deemed appropriate for this study.

Existing studies often have a relatively small sample size of individuals aged 50 and above [100,101]. However, online shopping is increasingly popular among older adults in China [7], and ADS may serve as a complementary tool for their shopping needs [102]. To gain a more comprehensive understanding of ADS adoption in this age group, this study increased the sample size of respondents aged 50 and above compared with previous studies. This approach aimed to provide a more accurate reflection of the consumer attitudes and adoption intentions of consumers across different age groups.

The survey instrument comprised 46 items encompassing four demographic parameters (gender, age, educational attainment, and monthly income) alongside 11 core constructs: perceived usefulness, ease of use, enjoyment, self-efficacy, participation, sustainability, risk perception, attitude, usage intention, personal innovativeness, and technology-related anxiety. Each construct was assessed via multiple indicators adapted from prior studies and rigorously validated to ensure methodological robustness. Responses were recorded using a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), with detailed measurement items and their origins tabulated in

Table 1. Survey item composition.

Measurement Items		References
Perceived Usefulness (PU)	Using an autonomous delivery service will be very useful to me.	[19,27,29]
	Using an autonomous delivery service will allow me to complete related tasks more efficiently.
	Using an autonomous delivery service will make deliveries faster.
	Using an autonomous delivery service will fulfill my needs.
Perceived Ease of Use (PEOU)	The process of using the autonomous delivery service will be simple and easy to understand.	[19,27,29]
	Using an autonomous delivery service will not require much mental effort.
	An autonomous delivery service will be easy to use.
	I will find it easy to learn how to use autonomous delivery services.
Perceived Enjoyment (PE)	I will feel great enjoyment when receiving packages through an autonomous delivery service.	[26,28,31]
	An autonomous delivery service will provide me with an enjoyable experience.
	An autonomous delivery service will offer a new and interesting experience.
Sustainability (SUS)	An autonomous delivery service is an environmentally friendly delivery method.	[38,39,40]
	An autonomous delivery service has low carbon dioxide emissions in the transportation process.
	An autonomous delivery service contributes to conserving natural resources.
	An autonomous delivery service can help reduce traffic congestion.
Customer Participation (CP)	I will actively participate in the autonomous delivery service process.	[43,50,103]
	I will collaborate with the service provider to complete necessary procedures (e.g., providing accurate addresses, setting delivery times).
	I will actively provide feedback to the service provider regarding any issues with the autonomous delivery service.
	I will make suggestions to the service provider for improvements or new features in the autonomous delivery service.
Self-efficacy (SE)	I am confident that I can easily operate an autonomous delivery service.	[61,64]
	I am confident that I can use an autonomous delivery service more effectively than others.
	I am confident in my ability to use an autonomous delivery service.
Perceived Risk (PR)	I am concerned that using an autonomous delivery service may involve certain risks.	[38,41]
	I am concerned that the safety of an autonomous delivery service may not be guaranteed.
	I am concerned about the risk of personal information leakage when using an autonomous delivery service.
Attitude (ATT)	I think using an autonomous delivery service is a wise choice.	[38,74,78]
	I think using an autonomous delivery service is desirable.
	I think using an autonomous delivery service is a good idea.
	I have a positive attitude toward using an autonomous delivery service.
	I think using an autonomous delivery service will be a great experience.
Intention to Use (ITU)	I am willing to choose an autonomous delivery service.	[31,74]
	I am more likely to use an autonomous delivery service in the future.
	I will receive products through an autonomous delivery service.
	If given the opportunity, I would like to try an autonomous delivery service.
Technology Anxiety (TA)	I tend to avoid autonomous delivery technology because I am not familiar with it.	[86,87]
	I feel anxious about using an autonomous delivery service.
	I worry about whether I can properly use an autonomous delivery service.
	I am anxious that I might make a mistake when using an autonomous delivery service and not receive my product.
Personal Innovativeness (PI)	I generally adopt new technologies more quickly than those around me.	[27,104]
	I am willing to try using new technologies and devices.
	I believe learning new technologies is an important personal skill.
	I think new technologies can significantly improve my efficiency and productivity.

. Linguistic accuracy was ensured through professional translation verification of all survey components.

Before full-scale data collection, a pre-test survey was conducted with 35 Chinese respondents to improve the readability and clarity of the questionnaire. After completing the survey, respondents provided feedback on items they found difficult to understand, overly long, or challenging to answer. The feedback indicated that some respondents had difficulty accurately understanding the concept of “autonomous delivery services”. To address this, a more detailed explanation was added at the beginning of the questionnaire, along with relevant images to aid visual understanding.

This research focused on Chinese residents aged ≥20 years, with data collection spanning 28 December 2024–15 January 2025. A total of 603 participants were initially recruited, from which 526 valid responses were retained after eliminating incomplete or inconsistent submissions, yielding a valid response rate of 87.23%. Respondent profiles detailing demographic distributions are comprehensively summarized in

Table 2. General Profile of Respondents.

Classification	Category	Number	%
Gender	Male	279	53.0
	Female	247	47.0
Age	20–below 30	129	24.5
	30–below 40	157	29.8
	40–below 50	118	22.4
	50 and above	122	23.2
Education	High school/below	73	13.9
	University/college graduate	380	72.2
	Postgraduate/above	73	13.9
Monthly income	Less than 410 USD	57	10.8
	410–820 USD	148	28.1
	820–1231 USD	178	33.8
	More than 1231 USD	143	27.2
Frequency of online shopping (including product purchases and food delivery)	Never	0	0
	Rarely	8	1.5
	A few times a year	31	5.9
	A few times a month	167	31.8
	A few times a week	264	50.1
	Every day	56	10.7

.

After data collection, all statistical analyses were performed using SPSS 26.0 and AMOS 24.0. To ensure the validity and reliability of the latent variables, confirmatory factor analysis (CFA) was performed on the final measurement model, which comprised 42 measurement items. Second, the structural model assessment adopted a two-stage analytical strategy. Initially, variable interrelationships were systematically analyzed. Subsequently, multi-group comparative analyses were performed to investigate the potential moderating effects of personal innovativeness and technology anxiety within the proposed framework.

4. Results of the Analysis

4.1. Confirmatory Factor Analysis (CFA)

The psychometric properties including standard item loadings, Cronbach’s α coefficients, average variance extracted (AVE), and composite reliability (CR) are summarized in

Table 3. Reliability and validity results.

Variable	Factor	Standard Item Loadings	Cronbach’s α	AVE	CR
PU	PU1	0.816	0.873	0.634	0.874
	PU2	0.763
	PU3	0.800
	PU4	0.806
PEOU	PEOU1	0.763	0.866	0.622	0.868
	PEOU2	0.844
	PEOU3	0.821
	PEOU4	0.723
PE	PE1	0.776	0.849	0.655	0.850
	PE2	0.843
	PE3	0.808
SUS	SUS1	0.803	0.860	0.606	0.860
	SUS2	0.768
	SUS3	0.772
	SUS4	0.771
CP	CP1	0.815	0.893	0.676	0.893
	CP2	0.811
	CP3	0.854
	CP4	0.809
SE	SE1	0.855	0.869	0.694	0.871
	SE2	0.880
	SE3	0.760
PR	PR1	0.844	0.901	0.752	0.901
	PR2	0.856
	PR3	0.901
ATT	ATT1	0.778	0.915	0.685	0.915
	ATT2	0.858
	ATT3	0.835
	ATT4	0.769
	ATT5	0.894
ITU	ITU1	0.824	0.892	0.675	0.892
	ITU2	0.836
	ITU3	0.826
	ITU4	0.800
TA	TA1	0.756	0.868	0.622	0.868
	TA2	0.815
	TA3	0.811
	TA4	0.771
PI	PI1	0.757	0.843	0.579	0.846
	PI2	0.777
	PI3	0.718
	PI4	0.790
Chi-square (χ2) = 935.656, df = 764, χ2/df = 1.225; p = 0.000, GFI = 0.924; CFI = 0.987; AGFI = 0.910; RMSEA = 0.021

. Following established psychometric standards, α values exceeding 0.7 were deemed acceptable for assessing scale reliability [105]. Internal consistency coefficients in this investigation ranged from 0.843 to 0.915, confirming adequate scale reliability.

Convergent validity assessment involved evaluating standardized factor loadings for all measurement items, each surpassing the 0.5 minimum threshold. CR coefficients demonstrated values between 0.846 and 0.916, exceeding the 0.7 benchmark, whereas AVE measurements spanned 0.579 to 0.752, all above the recommended 0.5 criterion. These findings collectively indicate the measurement model’s satisfactory convergent validity [106].

The structural equation model demonstrated adequate model fit (χ² = 935.656, degrees of freedom = 764, χ²/df = 1.225; p < 0.001, GFI = 0.924, CFI = 0.987, AGFI = 0.910, RMSEA = 0.021), with all metrics aligning with established psychometric benchmarks. Established criteria for model acceptability require χ²/df ratios below 3.0, GFI and AGFI values exceeding 0.8, CFI scores above 0.9, and RMSEA estimates under 0.08 [107,108].

Discriminant validity evaluation adhered to Fornell and Larcker’s [109] established criteria, which require that each latent construct’s AVE square root surpasses its correlation coefficients with other constructs. As documented in

Table 4. Discriminant Validity Results.

	PU	PEOU	PE	SUS	CP	SE	PR	ATT	ITU	TA	PI
PU	0.796
PEOU	0.425	0.789
PE	0.367	0.287	0.809
SUS	0.371	0.433	0.336	0.778
CP	0.282	0.372	0.309	0.329	0.822
SE	0.307	0.387	0.253	0.361	0.255	0.833
PR	−0.342	−0.337	−0.271	−0.309	−0.380	−0.238	0.867
ATT	0.437	0.531	0.324	0.441	0.393	0.407	−0.337	0.828
ITU	0.338	0.375	0.197	0.284	0.296	0.386	−0.293	0.696	0.821
TA	−0.229	−0.311	−0.150	−0.226	−0.270	−0.285	0.193	−0.506	−0.453	0.788
PI	0.204	0.253	0.130	0.258	0.188	0.211	−0.196	0.453	0.503	−0.440	0.760

Note: The square root of AVE is shown by the diagonal values.

, AVE-derived values exceed inter-construct correlations, confirming sufficient discriminant validity and demonstrating the measurement model’s lack of multicollinearity concerns.

4.2. Hypothesis Testing

The structural equation model exhibited satisfactory goodness-of-fit (χ² = 659.292, df = 498, χ²/df = 1.324; p < 0.001, GFI = 0.933, CFI = 0.985, AGFI = 0.920, RMSEA = 0.025), aligning with established psychometric benchmarks. Hypothesis testing outcomes are systematically detailed in

Table 5. Results of the direct standardized effect.

Path	Estimate	S.E.	C.R.	p	Path Result
H1: PU→ATT	0.155	0.046	3.234	0.001 **	Accepted
H2: PEOU→ATT	0.047	0.053	5.154	0.000 ***	Accepted
H3: PE→ATT	0.265	0.045	1.048	0.295	Rejected
H4: SUS→ATT	0.132	0.045	2.710	0.007 **	Accepted
H5: CP→ATT	0.131	0.040	2.883	0.004 **	Accepted
H6: SE→ATT	0.158	0.044	3.520	0.000 ***	Accepted
H7: PR→ATT	−0.059	0.042	−1.320	0.187	Rejected
H8: ATT→ITU	0.699	0.056	14.344	0.000 ***	Accepted
Chi-square = 659.292, df = 498, χ2/df = 1.324; p = 0.000, GFI = 0.933; CFI = 0.985; AGFI = 0.920; RMSEA = 0.025

** p < 0.01; *** p < 0.001.

.

Initial analysis of H1–H3 revealed that perceived usefulness (β = 0.155, p < 0.01) and perceived ease of use (β = 0.047, p < 0.001) exerted statistically significant positive influences on user attitudes, thereby validating H1 and H2. Contrarily, perceived enjoyment (β = 0.265, p > 0.05) failed to demonstrate a significant impact, resulting in H3’s rejection. These empirical findings suggest that utilitarian functionality and operational accessibility outweigh recreational aspects in ADS evaluations. Furthermore, sustainability demonstrated significant positive associations with attitude formation (β = 0.132, p < 0.01), corroborating H4 and confirming its critical role in attitudinal development.

Subsequent evaluation of H5–H6 investigating consumer trait impacts revealed that customer participation (β = 0.131, p < 0.01) and self-efficacy (β = 0.158, p < 0.001) exhibited significant positive correlations with attitudinal measures, supporting both hypotheses. This underscores the dual influence of interactive engagement and perceived competence alongside technical specifications in ADS assessments.

Contrary to initial hypotheses, perceived risk showed non-significant effects on attitude (β = −0.059, p > 0.05), necessitating H7’s rejection. Notably, attitudinal disposition strongly predicted usage intention (β = 0.699, p < 0.001), confirming H8 and underscoring attitude’s pivotal role in ADS adoption dynamics.

4.3. Multi-Group Analysis

This study investigated how personal innovativeness and technology anxiety moderate the relationship between attitude and intention to use, employing a multi-group structural equation modeling (SEM) analysis. Employing the grouping methodology established by Cheng et al. [110], participants were stratified into low- and high-TA cohorts using 3.324 (mean TA score) as the cutoff. Subsequent between-group comparisons employed a chi-square difference analysis between constrained and unconstrained models. The statistically significant disparity (Δχ² = 14.077, Δdf = 1) observed in model fit validates hypothesis H10, as detailed in

Table 6. Assessment of the Moderating Effect of Technology Anxiety.

Path	Δχ2, ∆df	Low (n = 281)		High (n = 245)		Result
		Estimate	C.R.	Estimate	C.R.
H9: ATT→ITU	Δχ2 (df = 1) =14.077 ***	0.789 ***	12.420	0.358 ***	4.520	Accepted

*** p < 0.001.

.

Attitude had a significant positive effect on intention to use in both groups, but this effect weakened as technology anxiety increased. Specifically, individuals with higher technology anxiety exhibited greater uncertainty and psychological resistance when encountering new technologies such as ADS. Despite having a positive attitude toward the service, their actual intention to use it decreased because of concerns about operating the technology, doubts about system reliability, and anxiety over unexpected issues. Conversely, individuals with lower technology anxiety were more proactive in translating their positive attitude into a higher intention to use. Therefore, technology anxiety serves as a moderating factor that weakens the conversion of attitude into intention to use.

Next, this study divided the sample into low and high personal innovativeness groups based on the mean value of personal innovativeness (3.709), applying the same method. A chi-square difference test was conducted to compare the restricted and unrestricted models, and the results showed a significant difference (Δχ² = 12.761, Δdf = 1). Therefore, hypothesis H9 was supported (

Table 7. Assessment of the Moderating Effect of personal innovativeness.

Path	Δχ2, ∆df	Low (n = 190)		High (n = 336)		Result
		Estimate	C.R.	Estimate	C.R.
H10: ATT→ITU	Δχ2 (df = 1) = 12.761 ***	0.362 ***	4.164	0.771 ***	12.397	Accepted

*** p < 0.001.

).

Analysis revealed differential attitudinal impacts on usage intention across innovativeness levels. While both cohorts demonstrated positive attitude–intention associations, the standardized path coefficient was more pronounced among highly innovative individuals (β = 0.762) relative to their less innovative counterparts (β = 0.636). This divergence aligns with established behavioral patterns wherein technology-progressive users exhibit greater receptivity to novel systems, facilitating accelerated adoption through enhanced experiential engagement. Conversely, technology-cautious users prioritize system dependability and operational stability during evaluation phases, which results in moderated attitude–behavior translation efficiency. These empirical patterns substantiate personal innovativeness’s moderating role in attitude–intention dynamics, with amplified effects observed in high-innovativeness populations.

As illustrated in Figure 2, H1, H2, H4, H5, and H6 were supported, which indicates that perceived usefulness, ease of use, sustainability, participation, and self-efficacy have significant positive effects on attitudes toward autonomous delivery services (ADS). H3 and H7 were not supported, which suggests that perceived enjoyment and perceived risk do not have significant effects on attitude. H8 was supported, which demonstrates that attitude significantly influences the intention to use ADS. H9 was supported as well, which indicates that technology anxiety negatively moderates the relationship between attitude and usage intention. H10 was also supported, meaning that personal innovativeness positively moderates the relationship between attitude and usage intention.

5. Discussion

In China, ADS based on drone technology and autonomous delivery robots have rapidly advanced in recent years because of their significant advantages over traditional delivery methods, including addressing environmental concerns, reducing costs, and alleviating traffic congestion. As ADS are increasingly recognized as a key delivery method for the future, understanding consumer adoption of this service is crucial. To this end, this study empirically tested the research hypotheses using 526 survey samples collected in China, yielding both theoretical and practical implications.

5.1. Theoretical Implications

This study explored key factors influencing consumer attitudes toward ADS and their intention to use these services, offering a distinct academic perspective compared with previous research. By expanding the theoretical foundation of ADS, it provides valuable insights and references for future research in related fields.

First, online shopping is becoming increasingly widespread among older adults in China [111], and ADS are expected to gradually replace conventional delivery methods [7]. However, previous studies have not sufficiently accounted for individuals aged 50 and above, limiting their ability to comprehensively capture attitudes and intentions across different age groups. To address this gap, this study expanded the sample proportion of consumers aged 50 and above to provide a more accurate understanding of ADS acceptance across all age groups. This contributes to a stronger theoretical foundation for future research.

Second, this study validated the effects of TAM factors, sustainability, and self-efficacy on consumer attitudes toward ADS. The findings reaffirm the applicability and validity of the TAM framework in ADS research, while also confirming that sustainability and self-efficacy play a crucial role in shaping consumer attitudes. This strengthens the theoretical basis for further studies on ADS adoption.

Third, in China, ADS typically require consumers to pick up their deliveries from designated locations [14]. In this process, consumers may perceive themselves as active participants in the service, which could influence their overall service experience [53]. To account for this, this study introduced “customer participation” as a new variable within the TAM framework and analyzed how active involvement in the service process impacts attitudes toward ADS. The findings suggest that greater customer participation leads to a more positive attitude toward ADS. This finding not only expands existing ADS research but also offers a novel theoretical perspective on how consumer attitudes are shaped in this context.

Finally, this investigation incorporated technology anxiety and personal innovativeness as moderators to examine their regulatory effects on the attitude–intention linkage within ADS adoption contexts. Analytical outcomes revealed a significant negative moderation by technology anxiety (β = −0.15, p < 0.05), which attenuates attitude’s predictive capacity for usage intention among anxiety-prone individuals. Conversely, personal innovativeness demonstrated positive moderation (β = 0.22, p < 0.01), with innovation-oriented consumers exhibiting enhanced attitude–behavior translation efficiency. These differential moderation patterns elucidate the psychological mechanisms underlying technology adoption decisions, advancing theoretical frameworks for understanding how individual traits interact with technological perceptions in behavioral modeling.

5.2. Practical Implications

ADS are gaining attention as an innovative technology in the distribution industry, contributing not only to operational efficiency and cost reduction for companies but also to environmental protection and urban traffic improvement [5]. This study analyzed the factors influencing consumer attitude and intention to use ADS and, on the basis of the findings, provides practical implications for companies to enhance adoption rates.

Initial analysis identified perceived usefulness and perceived ease of use as primary predictors shaping consumer attitudinal dispositions toward ADS. This observation corroborates Park et al.’s [27] empirical work investigating consumer adoption of self-service technologies within the fashion retail sector. Accordingly, delivery companies should optimize the efficiency of ADS and minimize waiting times, allowing consumers to experience faster and more reliable service. In other words, the greater the perceived convenience and benefits of ADS, the stronger the consumer preference for autonomous delivery. Moreover, companies should focus on streamlining operational procedures, optimizing user interfaces, and providing clear usage guidelines to reduce the learning curve for new users and make ADS operation more intuitive. These measures will lower adoption barriers, enhance usability, and facilitate consumer familiarity with the service. By improving these key aspects, companies can enhance the user experience, accelerate market penetration, and drive ADS diffusion.

However, perceived enjoyment was found to have no significant effect on attitude. While this contrasts with some prior studies, it is consistent with findings by Blut et al. [60] and Turan et al. [112] on SST and gamification tools. Previous research suggests that enjoyment significantly influences hedonic SSTs but has little impact on utilitarian SSTs [59]. Because the core value of ADS lies in their functional utility rather than entertainment, consumers prioritize speed, efficiency, and ease of use over enjoyment. While ADS incorporate innovative elements such as drones and robots, these novel experiences alone may not significantly influence consumer attitudes. In other words, functional benefits outweigh perceived enjoyment in consumer decision-making regarding ADS.

Second, consumer perceptions of sustainability had a significant positive effect on attitudes toward ADS, aligning with Klein and Popp [40] in their study on last-mile delivery. Across both developed and developing countries, interest in eco-friendly products and services is growing [39]. Consumers perceive that ADS contribute to environmental protection, resource conservation, and traffic reduction, leading to a more favorable attitude toward the service. This finding underscores the importance of sustainability in consumer decision-making when adopting ADS. Therefore, delivery companies should emphasize the eco-friendly aspects of ADS, particularly their low carbon footprint and sustainability advantages, in marketing and promotional efforts. By doing so, they can increase consumer awareness, enhance market acceptance, and position ADS as a sustainable delivery solution.

Third, customer participation exerts a pivotal influence on ADS-related attitude formation, consistent with prior empirical evidence regarding consumer engagement dynamics [53,56]. Within ADS operational frameworks, consumer roles transition from passive recipients to proactive collaborators actively shaping service delivery processes. Actively cooperating in deliveries, providing feedback, and suggesting improvements enhance the user experience and positively influence perceptions of ADS. Therefore, companies should view consumers not just as service recipients but also as “co-creators” and encourage participation through incentives such as rewards for feedback, membership programs, and delivery fee discounts. These measures can accelerate the adoption and market penetration of autonomous delivery technology.

Moreover, these findings align with SST research by Hsiao and Tang [63], confirming that consumer self-efficacy significantly influences positive attitudes toward ADS. The more confident consumers are in their ability to operate and use the technology, the more favorable their perception of ADS. To strengthen this, companies should optimize user training and guides, enhance service usability, and provide real-time technical support. These efforts will boost consumer autonomy and confidence, reinforcing both attitude and intention to use ADS.

Fourth, contrary to expectations, perceived risk did not significantly affect consumer attitudes toward ADS. While this finding differs from some previous studies, it partially aligns with findings by Hwang and Choe [69] and Mathew et al. [39] on ADS. Advances in autonomous delivery technologies, such as drones and delivery robots, have significantly improved service stability and safety. Additionally, as industry standards and regulations develop [113], consumer trust continues to grow. Increased awareness of ADS and the implementation of safety measures—such as system security, privacy protection, and compensation for lost items—may explain why perceived risk was not a decisive factor. This suggests that consumer attitudes toward ADS are influenced more by practical benefits than by concerns over risk.

Fifth, empirical validation confirmed the substantial positive association between consumer attitudes toward ADS and usage intention, mirroring conclusions from autonomous technology studies by Leong and Koay [74] and Edrisi and Ganjipour [41]. Specifically, attitudinal positivity exhibited a proportional correlation with service adoption propensity, which underscores attitude’s pivotal function in technological acceptance mechanisms. These findings emphasize the necessity of cultivating favorable consumer perceptions to enhance technology adoption trajectories. To achieve this, companies should emphasize the practicality, convenience, and sustainability of ADS. Furthermore, they should implement strategies such as market promotion, user experience optimization, reward systems for feedback, and incentives such as delivery fee discounts. These efforts will help drive ADS adoption and technological diffusion.

Sixth, technology anxiety negatively moderates the relationship between consumer attitude and intention to use ADS. In other words, the more unfamiliar consumers are with autonomous delivery technology, the more likely they are to avoid using it. Consumers may feel anxious about whether they can operate ADS properly on their own or worry about delivery failures due to operational mistakes. These concerns weaken the positive impact of attitude on intention to use, reducing the likelihood of adoption.

To increase consumer intention to use ADS, it is crucial to alleviate technology anxiety. Companies should optimize the user experience, provide clear operational guidelines, and enhance education and information delivery. For instance, they can increase familiarity with ADS by offering online tutorials on their operation and safety as well as expanding demonstrations and hands-on experiences. Furthermore, reducing technology anxiety through strategies such as improving delivery stability and enhancing transparency will further promote ADS adoption.

Finally, personal innovativeness positively moderates the relationship between consumer attitude and intention to use ADS. Consumers with high innovativeness adopt new technologies faster, viewing the ability to learn and master them as an important skill. Furthermore, they believe technology enhances their efficiency and productivity. To appeal to these consumers, companies should highlight the technological excellence of ADS and create an innovative atmosphere.

Conversely, consumers with low innovativeness tend to feel uncertain about new technologies and exhibit higher resistance. To address this, companies should actively share positive experiences and success stories from early adopters, emphasizing the stability and utility of ADS to build consumer trust. Moreover, offering ADS for free for a limited period can help consumers experience the service firsthand, reducing psychological resistance and increasing market acceptance. These efforts will contribute to broader ADS diffusion.

5.3. Limitations and Future Research

Although the present study reveals important findings, it has several limitations. First, it measured overall risk as a single indicator rather than segmenting different dimensions of perceived risk. This approach may introduce bias, as different types of risks—such as privacy risk and functional risk—may have distinct effects on consumer attitude and intention to use ADS. By aggregating all risks into a single dimension, the specific impact of each type on consumer decision-making may not be sufficiently captured. Future research should differentiate these risk dimensions and analyze their individual effects on consumer behavior in greater detail.

Second, this study did not examine the various dimensions of sustainability separately. Sustainability encompasses environmental, economic, and social aspects, each of which may influence consumer attitudes and intentions to use ADS differently. Future research should consider these sub-dimensions to better understand the role of sustainability in ADS adoption.

Thirdly, this study relies on a voluntary sampling method, and the sample size is relatively small, which may lead to selection bias. Future research should employ more random sampling techniques and expand the sample size to enhance the representativeness of the findings. Additionally, the sample in this study is primarily composed of individuals with high levels of education and income, who are more likely to hold positive attitudes toward online shopping. Furthermore, this study did not take regional differences into account, which may further exacerbate potential biases. Therefore, subsequent research should fully consider these factors.

Finally, ADS have only recently been introduced to the Chinese market, and consumer usage experience is still developing. As the market matures and user experience increases, future studies should examine how ADS influence continued intention to use, customer loyalty, and satisfaction over time.