Introduction

RUCP serve as designated parking spaces that not only alleviate parking pressure but also enhance environmental quality1. Their safety is particularly crucial in modern urban housing, encompassing not just structural stability, comprehensive facilities, and efficient traffic organisation, but also the prevention of security threats such as crime2. Parking safety is essential not only to prevent accidents and property damage but also to ensure a low-risk, comfortable parking environment. Research shows that garage safety is directly linked to RS and sense of RWB; a high-quality parking environment helps reduce travel-related anxiety and fosters a stronger sense of community belonging3.

Although relevant policies and standards already exist in China (such as the Residential Design Code and the Urban Residential Area Planning and Design Standards), in practice, there is still a lack of in-depth research into the relationship between PS and RWB. PS refers to individuals’ subjective assessment of potential risks, influenced by environmental, experiential, and personal factors4. In enclosed spaces such as underground car parks, design elements like lighting, ventilation, and wayfinding systems have a direct impact on PS5, while blind spots in surveillance and public security risks may lead to fear and avoidance behaviours6. Furthermore, the level of trust residents place in management also affects their sense of PS and psychological experience7. Therefore, exploring PS in RUCP and its impact on satisfaction and well-being not only contributes to improving safety design but also provides theoretical support for creating more human-centred urban living environments.

Existing research on RUCP has primarily focused on design optimisation8, safety management, and air quality9,10, with particular attention to fire safety, ventilation, smoke extraction, structure, and lighting systems, as well as their integration with architectural and auxiliary facilities. Chang and Choi8 highlight that lighting conditions, spatial layout, and safety management significantly affect residents’ psychological comfort. The application of natural ventilation and energy-efficient lighting technologies improves air quality while reducing energy consumption. High-quality lighting and ventilation systems are critical to ensuring safety and comfort in RUCP11,12. Dong et al.3 demonstrate that efficient spatial utilisation and layout design enhance garage safety and usability, while optimised space management, improved lighting and ventilation, and advanced surveillance systems reduce accidents and crime rates13,14. Wang et al.15 emphasise the importance of fire safety measures, including fire zones, automatic sprinkler systems, and robust firefighting infrastructure. Modern technologies, such as video surveillance, intrusion detection, and smart security systems, substantially enhance RUCP safety by preventing and mitigating risks16,17,18. Additionally, Dong et al.3 show that improved lighting and ventilation alleviate drivers’ anxiety and enhance their parking experience. Case studies by Lanza et al.19 underline the value of integrating intelligent technologies like video monitoring and automated parking in high-density urban areas to enhance safety and efficiency. Costa et al.20 argue that enforcing safety regulations and conducting regular training for residents and managers are pivotal for elevating safety standards. Finally, Chen et al.21 report that convenient parking facilities not only ease daily routines but also reduce the emotional stress associated with parking processes.

In summary, research on RUCP spans multiple disciplinary fields, including ergonomics, safety management, environmental psychology, and traffic engineering. Ergonomic studies suggest that creating a safe parking environment necessitates a thorough consideration of drivers’ psychological and behavioural characteristics, alongside scientifically designed layouts and access points to minimise accident risks22. Safety management theories emphasise that implementing robust regulations and management practices, such as surveillance systems and emergency response mechanisms, can significantly enhance parking safety23. Environmental psychology research has demonstrated that factors such as lighting, ventilation, and signage within parking facilities profoundly influence user behaviour and mental states; well-designed environments effectively improve comfort and PS3. Finally, traffic engineering provides essential technological solutions to ensure parking safety, including intelligent parking systems and vehicle detection technologies that optimise operational efficiency and reduce hazards17.

Analysis of existing literature indicates that findings on RUCP align well with the analytical framework of the SHEL model, which encompasses Software, Hardware, Environment, and Liveware. The SHEL model, a theoretical framework emphasising the interplay of multiple factors in complex systems, is frequently employed in risk analysis within aviation safety and other safety–critical domains24. In the context of parking facilities, hardware issues such as inadequate signage systems and limited parking space, environmental conditions like poor lighting and substandard air quality, and mismatches between residents’ lifestyle habits and facility design can collectively pose significant safety threats25,26. Adopting the SHEL model perspective enables a more systematic investigation into PS in RUCP, offering insights into how scientific planning, optimised design, and effective management can create safe and comfortable parking environments for residents. This integrative research approach holds substantial potential for enhancing RS and RWB.

Literature review and theoretical hypotheses

SHEL model

The SHEL model was first introduced by Professor Elwyn Edwards in 1972 and later visualised by Frank Hawkins in 197527,28. Widely applied in fields such as aviation and healthcare, the SHEL model facilitates the analysis of interactions between humans, equipment, and the environment27. In recent years, its scope has expanded to include the built environment and design domains, providing a framework to address safety issues in residential settings29. As a vital tool in ergonomics, the SHEL model is distinguished by its systematic and holistic approach, making it particularly suited for analysing and improving safety in complex systems30. In the context of RUCP and PS, the four elements of the SHEL model—Software, Hardware, Environment, and Liveware—offer a structured methodology to identify and evaluate potential risk factors and to develop targeted measures for improvement.

Specifically, the interaction of these four elements is shown in Fig. 1 and explained below:

Fig. 1
figure 1

(Modified from Hawkins,1987).

Framework of the SHEL model

Full size image

Liveware-Software (L-S)

In parking safety management, software elements include parking management systems, operating procedures and safety regulations31. The easy recognition of the licence plate recognition system and road gate system effectively helps drivers enter the parking quickly. Proper regulations and operating procedures can regulate the parking and circulation of vehicles and improve the operational efficiency and safety of the garage. Training of garage management and security personnel is important to ensure that they can operate the equipment correctly and handle emergencies.

Liveware-Hardware (L-H)

Hardware elements cover parking equipment, monitoring systems, ventilation and lighting systems. Modern parking equipment and intelligent monitoring systems can monitor the situation in the garage in real time, detect and deal with abnormalities in a timely manner, and improve overall safety32. Optimised ventilation systems and lighting facilities improve the air quality and visual environment of the garage, thereby reducing the likelihood of accidents3. In addition, the installation of fire fighting equipment such as fire alarms and automatic sprinkler systems ensures that the fire can be controlled in a timely manner in the event of fire to ensure the safety of personnel33.

Liveware-Environment (L-E)

The environmental element involves the physical and social environments. In RUCP, a good physical environment, such as appropriate lighting, temperature and noise control, can enhance users’ comfort and sense of security3. As for the social environment, a good community atmosphere and neighbourliness can help to reduce vandalism and criminal activities and enhance the overall sense of safety and well-being of residents34.

Liveware-Liveware (L-L)

Liveware factor involves the skills, knowledge, experience and psychological state of users and managers35. Drivers’ driving skills, parking habits and safety awareness directly affect the safety of garage use. Operational skills, emergency response capability and service attitude of management personnel (e.g., garage keepers and security guards) are important for the safe management of garages. Regular training and education can enhance the safety awareness and coping ability of garage users and management personnel to ensure that they can properly use the garage facilities and handle emergencies.

Research has shown that well-developed parking management systems and clear regulations significantly improve the operational efficiency and safety of car parks36. Modern hardware significantly reduce the incidence of parking accidents and theft13. Hardware enhancements directly improve the security of the parking environment. The physical and social environments also have a significant impact on PS37. Good lighting and ventilation design enhance users’ sense of security, while a good community atmosphere reduces vandalism3. These environmental factors directly enhance PS. The behaviours and skills of car park users and managers are important factors affecting parking safety. Studies have shown that improving the emergency response skills of managers through training can significantly reduce the incidence of parking safety accidents38. Based on these findings, the following hypothesis is proposed: the four elements of the SHEL model—Software, Hardware, Environment, and Liveware—exert a significant influence on PS.

  • H1: Software factors have a positive effect on PS.

  • H2: Hardware factors have a positive effect on PS

  • H3: Environmental factors have a positive effect on PS.

  • H4: Liveware factors have a positive effect on PS.

Human well-being is profoundly influenced by the physical and social characteristics of environments, including spatial design, ambient quality, and PS. RUCP, as the first indoor spaces residents encounter upon entering their communities, play a pivotal role in shaping psychological experiences39. A well-designed spatial environment significantly enhances RWB40. Physical environmental factors in underground garages, such as lighting, temperature, and air quality, directly impact psychological comfort and happiness11. Superior physical conditions help alleviate anxiety and unease, thereby elevating happiness levels41. Krefis et al.42 note that brightness, cleanliness, and aesthetic appeal in spaces foster positive psychological responses, thereby improving life satisfaction and happiness. Within RUCP, well-maintained lighting, cleanliness, and orderly layouts enhance spatial comfort43. According to Mehrabian and Russell’s44 stimulus–response model, perceived environmental safety reduces anxiety and tension, fostering positive emotional experiences. In RUCP, features such as surveillance cameras, emergency buttons, efficient circulation paths, and appropriate barriers enhance PS, which has a direct or indirect positive correlation with happiness45. Social psychology research further indicates that environmental ambiance influences not only immediate experiences but also long-term life satisfaction. Reduced noise pollution, improved air circulation, and intuitive spatial design lower residents’ stress levels, increasing overall satisfaction with their living environment. This satisfaction, mediated by a “psychological-environmental congruence,” ultimately boosts happiness46. Based on this, the following hypothesis is proposed:

  • H5: Environmental factors have a positive effect on RWB.

PS and RS

RS reflects residents’ overall evaluation of their living environment and quality of life, shaped by various factors including community environment, safety, and convenience47. In residential contexts, Gifford48 emphasised that PS directly influences residents’ perceptions of comfort and stability in their neighbourhoods, which are integral components of RS. Gholipour et al.49 further noted that urban planning features, such as adequate lighting, visibility, and clear pathways, enhance PS and, by extension, increase satisfaction with the community environment. In specific contexts like underground garages, PS assumes particular importance. The inherent characteristics of underground spaces, including poor lighting and enclosed areas, can provoke fear or anxiety, reducing PS50. Conversely, design, maintenance, and the provision of surveillance facilities significantly determine residents’ subjective sense of safety51. These factors influence residents’ usage experiences and psychological comfort in parking environments, which in turn affect overall RS. Research has demonstrated that when residents perceive parking environments as safe, their satisfaction with the overall residential environment improves significantly52. This satisfaction transcends physical security, encompassing assessments of community management, neighbourly relations, and convenience53. Wang54 found that robust parking management and surveillance systems effectively reduce crimes and accidents in parking facilities, thereby enhancing psychological safety and satisfaction. Similarly, Armitage13 highlighted that comprehensive parking facilities and safety measures mitigate negative emotions and conflicts arising from parking issues. Hyung55 revealed that secure parking environments reduce property losses and fear of crime, contributing to improved RS. Finally, Koçak Güngör and Terzi56 noted that secure parking facilities elevate overall community quality of life by alleviating neighbour conflicts and traffic congestion. These findings collectively establish PS as a critical variable influencing RS. Based on these insights, the following hypothesis is proposed:

  • H6: PS positively impacts RS

RS and RWB

RS, a critical psychological variable reflecting individuals’ overall perceptions and evaluations of their living environment, has been extensively studied in recent years. Its influence on RWB has been well-documented. On the one hand, RS is closely linked to the extent to which material and psychological needs are fulfilled. Andrews and Withey57 argue that the comfort and convenience of the residential environment are core dimensions of individuals’ overall perception of life quality. A comfortable and expectation-meeting living environment effectively reduces psychological stress and enhances RWB. On the other hand, Diener et al.58 highlight that RWB is not only associated with economic conditions but also significantly affected by the quality of one’s living environment. As a highly frequented daily space, the residential environment plays a pivotal role in shaping RWB. Research further reveals that both the physical attributes of specific environments and individuals’ subjective perceptions of those environments influence RWB. Porteous59 emphasises that physical characteristics of residential environments—such as tranquillity, superior safety, and positive neighbourly relations—substantially improve RS, which, in turn, enhances psychological states and indirectly boosts RWB. Additionally, PS in underground garages contributes to RS by influencing core dimensions of the living experience, such as convenience and security, thereby generating spillover effects on overall RWB. Psychological studies underscore that RS represents a concrete subdimension of overall life satisfaction, while RWB reflects a higher-level realisation and perception of life goals. The positive and stable correlation between the two is well-established60.

Drawing from these insights, it is reasonable to infer that enhancing RS, particularly through improving PS in RUCP, significantly strengthens psychological comfort and a sense of belonging, ultimately elevating RWB. Accordingly, this study proposes the following hypothesis:

  • H7: RS positively influences RWB.

PS and RWB

The relationship between PS and RWB is a central concern across psychology, sociology, and environmental behaviour studies. Research indicates that individuals’ PS significantly influences their mental health and subjective well-being58. In residential contexts, PS extends beyond subjective perceptions of crime threats, encompassing spatial design, social interactions, and environmental management61. Residents in high-PS communities are more likely to experience elevated RWB, as PS reduces psychological stress and social anxiety, fostering a sense of control over their environment62. Particularly in public spaces such as underground garages, which are enclosed or semi-enclosed, PS plays a crucial role in residents’ psychological comfort and life satisfaction63. When individuals perceive an environment as safe, they are more inclined to engage in social activities, building a stronger sense of community belonging and social support networks—key contributors to happiness64. Conversely, perceptions of insecurity can lead to social isolation, reduced activity levels, and diminished RWB65. Given the narrow spaces, poor lighting, and high privacy levels of underground garages, they are often perceived as high-risk environments. However, measures such as optimising lighting and enhancing spatial layout can significantly improve PS, thereby boosting satisfaction with the overall residential environment and RWB66. Furthermore, the impact of PS on RWB is not limited to its influence on RS; PS also directly alleviates anxiety and stress, further enhancing subjective well-being67. Based on this evidence, the following hypothesis is proposed:

  • H8: PS positively influences RWB.

Theoretical model construction

Drawing on theoretical support from existing literature, this study develops a conceptual model based on the SHEL framework to examine the impact of PS on RS and RWB (as illustrated in Fig. 2). As a critical component of the residential environment, PS has been shown to significantly enhance RS while exerting a positive influence on RWB. This mechanism has been validated in numerous studies and demonstrates remarkable consistency across diverse cultural contexts, providing a robust theoretical foundation for this research.

Fig. 2
figure 2

Conceptual model.

Full size image

This model integrates the SHEL dimensions—Software, Hardware, Environment, and Liveware—to comprehensively analyse the interplay between PS, RS, and RWB. The framework posits that improving PS through targeted interventions in parking safety can directly or indirectly affect RS and ultimately contribute to higher RWB. By consolidating multidisciplinary insights, this theoretical model offers a systematic approach to understanding the intricate relationships within the residential environment.

Research methods

Research subjects and questionnaires

This study selects the main urban area of Lianyungang as its primary research focus. The urban characteristics and current development status of Lianyungang closely mirror those of many small- and medium-sized cities in China, rendering the findings and recommendations highly representative and broadly applicable. Consequently, the results provide valuable insights for the design and management of RUCP in similar cities. The survey scope encompasses mid- to high-rise residential areas in the Haizhou, Ganyu, and Lianyun districts of Lianyungang, ensuring data collection is both comprehensive and diverse.

The questionnaire employed a five-point Likert scale68 to quantitatively assess residents’ perceptions and evaluations of various safety-related factors. The questionnaire consisted of two sections: the first gathered demographic information, while the second addressed seven core variables, namely software factors, hardware factors, environmental factors, liveware factors, PS, RS, and RWB. Each variable was assessed using three items to ensure both comprehensiveness and reliability. Through a scientifically designed and pre-tested questionnaire, the study provides a realistic reflection of the perceptions and experiences of residents in Lianyungang in their everyday lives. Moreover, it offers valuable data support for research and urban management practices in other comparable cities, thereby enhancing the applicability and generalisability of the research findings.

To ensure the scientific rigour and validity of the questionnaire, a small-scale pretest was conducted prior to its official dissemination. Feedback from the pretest was used to refine and optimise the questionnaire to better align with respondents’ comprehension and real-life contexts. The questionnaire was designed and implemented via the “Sojump” platform, with data collected through two methods: face-to-face QR code scanning and distribution via WeChat groups. The survey period spanned from 15 July to 31 August 2024, yielding a total of 423 responses, of which 282 were valid, resulting in an effective response rate of 66.67%. All respondents were required to read and sign an informed consent form before completing the questionnaire, ensuring that the research process adhered to ethical standards. The sample statistics, summarised in Table 1, provide a robust foundation for analysing the factors influencing PS in RUCP. They also reveal significant differences in parking habits and attitudes across demographic groups, offering crucial support for subsequent in-depth analysis and research.

Table 1 Descriptive statistical analysis.
Full size table

Sources of scales for the measurement of each variable

In order to ensure the reliability and validity of the scales in this study, validated and well-established scales were used for all variable measurements. These scales have been tested in several studies have high reliability and validity, and can accurately measure the relevant variables. Based on the specific needs of this study, the content of the questions was appropriately adjusted and optimised to improve the comprehensibility of the questions and to ensure that respondents were able to accurately understand and respond to them. These adjustments not only retained the core measurement function of the original scale but also enhanced its applicability in specific research situations. Please refer to “Appendix 1” for detailed information and specific content of the questionnaire items for further reference and analysis. “Appendix 2” presents the mean, standard deviation, skewness, and kurtosis for each measurement item, reflecting the central tendency and dispersion of the data distribution. For most variables, the absolute values of skewness are less than 1, and the kurtosis values fall within the conventional range (± 2), indicating that the data approximate a normal distribution and are therefore suitable for structural equation modelling analysis.

Reliability and validity test of the scale

In this study, the internal consistency of the scales used was assessed by calculating Cronbach’s α coefficient to ensure the reliability of the scales. The internal consistency of the 21 question items was analysed using SPSS 26.0 software, and the results showed that Cronbach’s α values for each variable exceeded 0.7, indicating that the scale had high reliability. This result indicates that the scale used has strong reliability in measuring each variable, i.e., the scale is able to maintain a consistent measurement across different question items. Therefore, the results of the study have high reliability and internal consistency, and can effectively reflect the true opinions and feelings of the respondents. The detailed results of the reliability analysis are shown in Table 1.

In terms of validity testing, this paper focuses on assessing convergent validity and discriminant validity. Convergent validity is tested by calculating two indicators, the constitutive reliability (CR) and average explained variance (AVE) of the latent variables. The CR value reflects the internal consistency of the constructs, and Fornell et al. suggest that the CR value should be higher than 0.6 to indicate strong consistency, while the AVE measures the explanatory power of each measured variable to account for the variance of its corresponding latent variable, with a suggested value of 0.5 or higher69. As can be seen from Table 2, the CR and AVE values for each latent variable in this paper exceed the recommended standards, showing that the scale has good convergent validity.

Table 2 Reliability tests of the dimensional scales.
Full size table

In addition, the model was also considered to have good convergent validity when the factor loadings of each measure were greater than 0.5. According to the data in Table 1, the standardized factor loadings between each measure and its latent variable are greater than 0.5, further validating the convergent validity of the model. These results indicate that the scale used in this paper can effectively reflect respondents’ real responses when measuring each latent variable and has strong explanatory power in theoretical construction, ensuring the scientificity and validity of the study.

As suggested by Fornell et al.69 tests of discriminant validity can be assessed by comparing the AVE square root of each construct with the correlation coefficient between the constructs. Specifically, when the square root of the AVE for a construct is greater than the correlation coefficient between it and the other constructs, it indicates that the construct has good discriminant validity. Table 3 shows that the square root of AVE for each latent variable is higher than their correlation coefficients with other latent variables, indicating that the instruments measuring these constructs have good discriminant validity.

Table 3 Test of discriminant validity of latent variables.
Full size table

Model fit and hypothesis testing

Model Fit

Based on ensuring good reliability and validity of each research variable, this paper tested the fitness of the hypothesized model to the actual data using the AMOS 24.0 tool. To assess the fitness of the models, this paper relied on the fitness indicators recommended by several scholars, which are shown in Table 4. Given that the CMIN is highly sensitive to the sample size 70, and that almost all models may be judged unfit when the sample size is too large, this paper therefore used the CMIN/DF as an alternative indicator. According to Bollen’s suggestion, a model is considered to have a good fit when the CMIN/DF value is less than or equal to 371. Table 3 shows that all the indicators of the model in this paper meet or exceed the corresponding critical values, indicating that the model has a good fit. The structural equation model verification is illustrated in Fig. 3.

Table 4 Key indicators for model fit testing.
Full size table
Fig. 3
figure 3

Structural equation model validation diagram.

Full size image

Hypothesis testing

In testing the fit of the hypothetical model to the survey data using AMOS 24.0 software, the paper also obtained standardized path coefficients between the measured variables in the structural equation model and empirically tested each of the research hypotheses. Based on the results of the analyses in Table 5, the following conclusions are drawn: first, software factors have a significant positive impact on PS, thus supporting hypothesis H1. In addition, hardware factors also show a significant positive impact on PS, validating hypothesis H2. Although environmental factors do not have a significant impact on PS (not supporting hypothesis H3), they have a significant positive impact on RWB, supporting hypothesis H5. Meanwhile, the positive influence of liveware factors on PS is significant, which verifies hypothesis H4. PS has a significant positive influence on RS, which supports hypothesis H6, while RS also has a significant positive influence on RWB, which supports hypothesis H7. However, the influence of PS on RWB does not reach a significant level, and therefore, hypothesis H8 is not supported.

Table 5 Research hypothesis testing.
Full size table

Conclusions and limitations

Conclusions of the study

RUCP, as essential components of housing developments, serve as the initial indoor space residents encounter when entering their communities by vehicle. Influenced by standards for architectural, structural, and fire safety design, existing research has predominantly focused on structural safety, fire prevention, and air quality in underground garages. However, studies examining the impact of PS on RS and RWB remain limited. Furthermore, the SHEL model—a critical tool for analysing interactions among software, hardware, environment, and Liveware factors—has yet to be fully validated in the context of residential underground garages. This study adopts the SHEL model, using underground garages in mid- to high-rise residential areas in Lianyungang as the research setting, to explore the mechanisms through which PS influences RS and RWB. Through empirical research, this study provides a comprehensive analysis of the effects of PS in RUCP on residents, with the following theoretical contributions:

Validation of the SHEL Model: The SHEL framework demonstrates robust explanatory power in analysing PS influence on RS and RWB. While previous studies explored factors affecting parking safety, this is the first to systematically examine the interplay among PS, RS, and RWB using this model. The findings confirm direct impacts of software, hardware, and liveware on PS and highlight the indirect pathway through RS to RWB.

Firstly, the SHEL model demonstrates strong explanatory power in elucidating the relationship between PS, RS, and RWB in RUCP. While prior research has explored various factors influencing RUCP safety, this study marks the first application of the SHEL model to assess the relationships between PS, RS, and RWB. Overall, the software, hardware, and liveware within RUCP have a direct impact on PS. Furthermore, RS positively influences RWB, and although environmental factors in RUCP contribute to RWB, they do not directly affect PS.

Secondly, software factors exert a significant positive impact on PS, with a path coefficient of 0.270 (p < 0.05). By mitigating individual uncertainty and operational confusion, software factors improve residents’ perceived control over the environment. For instance, precise guidance on parking space allocation and streamlined vehicle entry and exit procedures can significantly reduce users’ psychological stress when navigating RUCP. Moreover, the higher path coefficient of software factors relative to other variables underscores their prominent influence, suggesting that enhancements in this domain can yield rapid and effective results.

Thirdly, hardware factors also exhibit a significant positive impact on PS, with a path coefficient of 0.268 (p < 0.05). This underscores the pivotal role of optimised hardware design in enhancing residents’ sense of security. Measures such as the strategic arrangement of lighting systems, clear signage for emergency exits, and the deployment of high-quality surveillance equipment can significantly alleviate residents’ feelings of insecurity in RUCP. Given that RUCP environments are often perceived as enclosed, concealed spaces fraught with potential risks, inadequacies in hardware facilities may exacerbate psychological distress.

Fourthly, although the research hypothesis proposed a positive relationship between environmental factors and PS (H3), the empirical findings revealed that this direct effect was not statistically significant (path coefficient = 0.018, p > 0.05). This result stands in stark contrast to the significant effects observed for software, hardware, and liveware factors. Two possible explanations may account for this finding. First, the influence of environmental factors on PS may be largely indirect, operating through mediating variables such as user experience, emotional state, or trust in community management. Second, respondents’ perceptions of environmental aspects within underground car parks may be more closely associated with evaluations of spatial comfort or aesthetics rather than with direct considerations of safety risks. This phenomenon is also reflected in existing literature; for example, Baek et al.39 suggest that physical environmental features often influence feelings of safety through complex cognitive appraisal processes, and that the strength and nature of this influence may be moderated by factors such as spatial enclosure, lighting ambience, and individual psychological traits. In contrast, H5 was supported, indicating that environmental factors have a significant positive effect on RWB (path coefficient = 0.197, p < 0.05). This suggests that even if environmental elements do not directly enhance PS, they nonetheless play a meaningful role in supporting RWB. A high-quality spatial environment—characterised by cleanliness, good ventilation, low noise levels, and pleasant lighting—can directly enhance residents’ emotional experiences and sense of belonging. As an extension of the residential space, the environmental quality of RUCP is perceived by residents as part of the overall living experience. Its comfort and design appeal may contribute more to well-being than to PS. Accordingly, the findings of this study suggest that environmental factors may contribute to subjective well-being primarily by enhancing the overall ambience of daily life rather than by directly increasing feelings of safety.

Fifthly, liveware factors exhibit a significant positive influence on PS, with a path coefficient of 0.283 (p < 0.05), ranking as the most impactful among all independent variables. This finding underscores the central role of interpersonal interactions in shaping perceptions of environmental safety. Within the context of RUCP, this influence may manifest through the professionalism and service attitude of garage management personnel, as well as a harmonious atmosphere among residents. For example, efficient and courteous behaviour from staff can significantly enhance residents’ trust, while mutual respect and cooperation among residents further reduce conflict and feelings of insecurity.

Sixthly, PS serves as a core variable, exerting both direct and indirect impacts on RS and RWB. PS has a strongly positive effect on RS, with a path coefficient of 0.375 (p < 0.01). A heightened sense of safety in RUCP directly enhances RS with their living environment. When PS are high, residents are more likely to generalise this positive experience to their overall perception of the community, thereby increasing RS. This outcome aligns with the “safety first” principle in housing satisfaction theory. RS significantly influences RWB, with a path coefficient of 0.225 (p < 0.05). This suggests that higher satisfaction with the residential environment corresponds to greater RWB. As a critical mediator, RS connects PS with RWB, offering actionable policy recommendations for the design of community environments. The direct effect of PS on RWB is less pronounced, with a lower path coefficient of 0.098 (p < 0.05). This result implies that the primary mechanism through which PS impacts RWB is mediated by RS. In other words, improvements in PS contribute to enhanced living experiences, which subsequently elevate RWB. This finding aligns with environmental psychology research, which suggests that PS functions as a foundational variable with a relatively weak direct link to RWB72.

In summary, this study is the first to systematically introduce the SHEL model into the investigation of PS within RUCP, demonstrating its applicability and effectiveness in explaining the relationships among PS, RS, and RWB. The research not only expands the application scope of the SHEL model but also addresses a theoretical gap in the study of safety perception within residential built environments. The findings provide empirical support for the optimisation of parking management and facilities in residential areas, highlighting the crucial role of parking management systems, well-developed hardware infrastructure, and driver safety education in enhancing residents’ well-being. Accordingly, it is recommended that urban planners and community managers adopt the systematic perspective of the SHEL model in the design and operation of residential areas. By integrating human factors engineering with environmental design principles, it is possible to improve the user experience of underground car parks—thus promoting resident well-being at the micro level and contributing to greater urban liveability at the macro level.

Research limitations and future research directions

Although this study strictly followed the standard procedures of questionnaires and empirical research, the following limitations exist. Firstly, there are deficiencies in the geographical and representativeness of the sample. The data of the study mainly came from residents of the main urban area of Lianyungang, so the findings may not be fully generalised to other regions with different cultural backgrounds or socio-economic conditions. Second, the subjectivity of variable measurement may affect the objectivity of the results. This study relied on questionnaires to collect data based on respondents’ self-reports, which may lead to measurements being affected by subjective perceptions and social expectation bias, which in turn may affect the accuracy of the data. Again, there are limitations in the research design. Although this study adopted a cross-sectional design to reveal the correlation between variables, the causal path between parking safety and residential satisfaction and well-being has not been verified through longitudinal studies or experimental designs. Finally, the simplified treatment of the model may have overlooked some key factors. Although the SHEL model provides a theoretical framework for this study, it may oversimplify the actual situation by not fully considering potential influences such as community atmosphere and government policies, which may also significantly affect parking safety, residential satisfaction and well-being.

To overcome the above limitations and further deepen the understanding of the field, future studies should consider expanding the scope and diversity of the sample to cover areas with different socio-economic backgrounds, cultural characteristics and levels of urban development so as to enhance the external validity of the findings. In addition, the introduction of more objective measures, such as actual parking accident data and behavioural observation records, can reduce the impact of bias due to self-reported data, thus improving the reliability and accuracy of the study results. Further studies should also adopt a longitudinal design or experimental approach to track the dynamic changes in residents’ parking safety experience and their residential satisfaction and well-being in order to verify the causal relationship between the variables. In addition, future research should consider more potential variables and mediating effects, such as community interaction and government policy support. Continuing and expanding research in these directions will help to understand the complex relationship between residential parking safety and residents’ quality of life more deeply and provide more scientific guidance for urban planning and residential management.