Residents’ Acceptance of Shared Autonomous Vehicles (SAVs) and Its Impact on Community Parking Demand Under Urban Regeneration: The Case of the Qintai Community in Wuhan, China

Abstract

Rapid urbanization and limited land resources have intensified parking shortages in China’s core and old urban districts, highlighting the tension between parking supply and public space. This study investigates the staged impacts of shared autonomous vehicles (SAVs) on private car ownership and parking demand within the context of urban renewal. Using a case study of Qintai Community in Wuhan, we combined resident surveys (135 valid samples), on-site parking facility assessments, and demand forecasting models to evaluate changes in parking requirements across different timeframes. Results indicate that SAVs can substantially reduce private car ownership and reshape parking demand structures, with short-term transitional pressures followed by long-term demand contractions. Furthermore, SAV adoption offers opportunities to reallocate parking land for multifunctional urban uses, alleviating land-use conflicts in high-density neighborhoods. The findings contribute to a dynamic framework for staged parking optimization, integrating technological innovation with community-level urban renewal strategies. This study underscores the importance of linking residents’ behavioral shifts with infrastructure adaptation, providing evidence-based guidance for sustainable urban transport and space management.

1. Introduction

With the acceleration of urbanization, the intensity of land use in urban cores and old urban areas has been increasing, and outstanding problems such as reduced environmental quality, limited access to public services, and aging infrastructures are emerging one after another [1]. Urban renewal is viewed as a crucial means to enhance the functionality and vitality of historic urban areas. In response to these problems and the challenges of sustainable development, the Chinese government has put forward an urban renewal strategy that focuses on improving the institutional mechanism, improving the quality of the environment and services, promoting economic and cultural development, and building a high-quality living space [2].

Insufficient parking space due to land constraints is a key pain point in the urban renewal process [3]. The competition between parking spaces and public spaces is becoming increasingly apparent. Excessive parking space not only compresses the space for residents’ leisure activities, but also weakens the overall quality of the public environment in the community [4,5]. The contradiction between supply and demand between increasing car ownership and limited parking space may have a negative impact on urban transportation and space utilization [6]. In the face of the long-term pressure on the parking resources brought about by the continuous growth of motor vehicle ownership, especially the rigidity of the space caused by the high proportion of private cars, China’s parking problem shows the trend of “total volume tends to slow down, the structure is still tight, and the management improves its efficiency”, which is somewhat typical.

Currently, international studies suggest that alleviating the shortage of parking space supply and reducing parking demand are the two most effective ways to solve the challenges [4,7]. Supply-side measures, such as constructing multi-story or underground parking facilities and promoting shared parking, can expand capacity to some extent; however, they are costly, land-intensive, and may inadvertently induce additional private car use [8,9]. Management-oriented approaches, including the deployment of smart parking systems to provide real-time information, dynamic pricing mechanisms, and stricter enforcement, can improve efficiency, yet they remain constrained by technological coverage, public acceptance, and equity concerns [10,11,12]. On the demand side, strategies such as enhancing public transport, carpooling, and intermodal transfers are intended to shift travel away from private cars toward more sustainable modes, but their effectiveness is limited by behavioral inertia and insufficient alternatives [13]. This highlights the urgent need to explore emerging mobility solutions that may fundamentally reshape parking demand patterns.

In the field of urban transportation, research on the relationship between travel modes and parking demand has accumulated a certain body of work, providing a theoretical foundation and practical reference for understanding parking supply and demand relationships in traditional travel scenarios. However, with the emergence of new technologies and travel paradigms, parking research is shifting from traditional supply-demand regulation toward more transformative directions. In particular, with the rise of new travel modes combining the sharing economy and autonomous driving technologies, the function of parking facilities has shifted from “serving individual drivers” to “supporting shared mobility networks.” At the same time, the travel mindset of residents has also changed, from “emphasizing vehicle ownership and autonomous driving” to “focusing on shared services, cost efficiency, and trust in technology.”

It is important to note that existing research is largely based on traditional scenarios built around “individual vehicle services” and “vehicle ownership orientation,” which fail to accommodate the core characteristics of the new scenarios, such as “supporting shared networks” and “service and efficiency orientation.” This makes it difficult to accurately assess whether these shifts can effectively alleviate urban parking problems from the demand side and fails to clearly define the future development direction of parking spaces. As a result, the research system has increasingly shown signs of insufficient adaptability, with significant gaps in three key areas.

Firstly, there is a lack of quantitative analysis of the relationship between new travel modes and parking demand. Most current research remains at the qualitative level, for example, suggesting that “new travel modes may have a suppressive effect on parking demand,” but failing to clarify the quantitative relationship between different types of new travel modes and changes in parking demand [14]. Secondly, there is a lack of exploration into the differentiated impacts on small-scale spatial units (such as old communities in the central urban area) with significant variations in land use types, population density, and the level of transportation infrastructure [15]. Finally, the research on the underlying mechanisms by which new travel modes reshape parking demand remains insufficient. Existing literature has not clearly disaggregated the internal logic and pathways through which new travel modes influence parking demand.

The development of Artificial Intelligence makes SAVs become a major driving force for the progress of urban intelligent transportation [14,15]. In recent years, the smart vehicle sector has been growing rapidly in the Internet of Things (IoT) market, and with the development of technology and policy inclination, the continuous increase in the number of AVs (autonomous vehicles) has become a major trend [14]. The emergence of SAVs not only implies the birth of a new mode of transportation but also has a great impact on the future of urban morphology, the environment, and the sustainability of society [16,17,18].

The ripple model has been used to conceptualize the potential sequential effects of autonomous driving on multiple aspects of mobility and society [19].

The first is the diversification of the impacts of autonomous driving technologies on travel costs and travel choices. Although the acquisition cost of AVs remains high at present, it is expected to decline substantially with large-scale production and technological advancements. At the same time, improvements in comfort, safety, and the potential for multitasking are likely to alter the perceived value of travel time, thereby influencing the Gross Travel Cost (GTC) [20,21,22,23]. In the short term, cost reductions and road capacity improvements may stimulate more motorized trips, increase total vehicle miles traveled (VMT), and lead to a decrease in the ratio of public transportation to non-motorized trips; however, under high-cost or differential pricing scenarios, shared automated driving may curb VMT growth while encouraging the adoption of public transport and active travel modes such as walking [24,25].

The second are the impacts on vehicle ownership and sharing, location choice and land use, and transportation infrastructure. The proliferation of SAVs could drive carpooling and vehicle sharing by lowering operating costs, increasing flexibility, resulting in significant reductions in vehicle ownership and parking demand [26]. At the regional level, they can improve accessibility, reduce travel costs, promote the spread of housing and employment to suburbs and even rural areas, and change urban form and land use; at the local level, they can free up on-street and off-street parking space, which can be converted into public space or other transportation-use land [27]. In terms of transportation infrastructure, increased road capacity and decreased parking demand may reduce the need for new roads and parking facilities, but increased travel demand may offset the capacity benefits; at the same time, the mode of operation, ownership, and spatial allocation of infrastructure may be structurally adjusted in response to the rising proportion of fully autonomous vehicles [28,29].

The third is the broader societal impacts, including energy consumption, air pollution, safety, social equity, economy, and public health. Autonomous driving can improve energy efficiency and reduce emissions by easing congestion and optimizing driving, as well as reducing accident rates, improving travel opportunities for disadvantaged groups, and economically reducing transportation costs and improving efficiency [16,30,31]. However, its diffusion still faces challenges such as high costs, low penetration rates, and cybersecurity risks; and if it leads to a reduction in public transportation and an increase in the number of trips made, it may also increase social inequality, impact employment in related industries, and bring health concerns such as reduced physical activity [32,33].

In China, the development and application of shared autonomous vehicles (SAVs) are strongly supported by a comprehensive policy framework and strategic planning. Current legislation and pilot programs for L3 and above autonomous driving permit both commercial and non-commercial activities using autonomous vehicles, while regulatory measures specific to intelligent connected vehicles at L3 and higher levels are being introduced [34,35]. Through the coordination of national and local policies, and by leveraging pilot cities as experimental grounds, China has actively facilitated the rapid growth of shared mobility models and accelerated the large-scale adoption of SAVs [36].

At the local level, numerous cities have launched pilot programs for autonomous taxis, shuttle services, and applications in closed environments, while also introducing local regulations addressing parking management, data security, and liability allocation. As one of China’s pilot cities for autonomous driving, Wuhan has witnessed rapid development in shared autonomous vehicles (SAVs) [37]. The “Apollo Go” Robotaxi service has deployed more than 400 vehicles, with steadily expanding coverage, increasing user adoption, and rising penetration. By mid-June 2024, the service had completed over 1.58 million ride orders, serving nearly 1.98 million passengers, with total testing mileage surpassing 100 million kilometers and the longest single trip exceeding 95 km [38]. Its operational area now encompasses the Wuhan economic and technological development zone and surrounding regions, extending to major commercial districts, residential areas, and universities, demonstrating the rapid diffusion and profound impact of autonomous mobility in the city.

Both internationally and domestically, the value of promoting SAV technology is significant. This technology, which integrates interconnection, sharing, and automation, will have a far-reaching impact on the urban public parking space [16,17,39,40,41].

Shared Autonomous Vehicles (SAVs), as an emerging travel mode integrating connectivity, sharing, and automation, are expected to offer a breakthrough solution to urban parking challenges by improving parking efficiency and reducing vehicle ownership. However, their effectiveness still requires further validation considering residents’ acceptance and stage-specific characteristics [42,43,44,45,46,47,48].

Given that traditional approaches fail to address the root causes of parking shortages in old residential communities, this study takes a typical Chinese community (Qintai Community in Wuhan) as a case study. It focuses on the core question—whether SAVs can alleviate community parking pressure in stages by influencing residents’ travel behavior—and verifies their effects through questionnaire surveys and model-based calculations, while proposing corresponding strategies. The research results can provide a dynamic decision-making basis for urban transportation planning, help solve the parking tension problem in phases, and promote the efficient use of urban space resources.

2. Theoretical Framework

2.1. Systematic Impact of SAV on Urban Transportation in China

As the world’s largest automobile market and one of the countries with the fastest urbanization process, China’s megacities have particularly prominent traffic congestion and parking conflicts, and are unique in terms of infrastructure construction standards, policy implementation efficiency, and technological landing scenarios, etc. Moreover, the application and popularization of SAV in China has already occupied an important position in the world, and it can provide a more adaptable reference paradigm for cities at different stages of development around the globe. SAV will profoundly reconstruct the road utilization, parking space, and infrastructure system of Chinese cities, and its impact will be deeply coupled with China’s “high-density, mixed land use” urban characteristics, presenting a dual orientation of efficiency improvement and functional transformation.

From an international perspective, the impact of SAVs on urban transportation involves complex trade-offs between technological diffusion and social efficiency. The dynamic vehicle scheduling and trajectory coordination enabled by widespread SAV adoption resonate with China’s context: megacities like Beijing and Shanghai could alleviate peak-hour congestion without expanding road networks. Concurrently, SAVs’ shared attributes will alter travel intensity by restricting private driving through differentiated access rights [44,49]. Consequently, road infrastructure must evolve into “intelligent sensing spaces” to support SAVs’ real-time communication needs and optimize road resource allocation [46,50]. SAV adoption can reconfigure parking spaces to resolve parking conflicts in Chinese cities: dedicated parking lots no longer require reserved spaces for drivers, while high utilization and shared attributes significantly reduce overall demand. Parking spaces can migrate to urban peripheries, freeing land for conversion into public spaces—aligning with “urban renewal” policy directions. The proliferation of SAVs will also drive the evolution of parking management toward intelligent and unmanned systems. This will enhance operational efficiency, reduce labor costs, and shift facility functions from “serving individual drivers” to “supporting shared networks,” enabling infrastructure transformation [41,46,51,52]. This is particularly crucial in Chinese cities facing land resource constraints.

2.2. Impact of SAV on Residents’ Willingness to Purchase Cars

Vehicle ownership reflects urban parking demand, and residents’ car-purchasing willingness stems from the dynamic balance of economic capacity, policy environment, usage needs, and other factors [53,54,55].

While the popularization of SAVs lowers the necessity of private car ownership, as their shared, pay-as-you-go, on-demand dispatch mode reduces users’ fixed costs and vehicle idling rates, increases utilization [26,29,43]. SAVs change travel costs and convenience compared to traditional modes, improved comfort and safety, and can self-park remotely to alleviate parking pressure [56,57,58].

2.3. Changes in the Supply and Demand of Urban Parking Spaces in the Context of SAVs

On the demand side, the popularization of SAV may change the urban parking demand structure through multiple mechanisms. First, in terms of vehicle ownership, the substitution effect of SAVs for private cars can significantly reduce the overall demand for parking spaces, especially in areas where parking resources are tight and the cost of purchasing a car is high, which is more likely to inhibit the willingness of residents to purchase a car [29]. Secondly, in terms of parking behavior, SAVs have automatic pick-up and drop-off and automatic parking functions, and can drive away to low-cost parking lots in the periphery after picking up and dropping off passengers in the core area, or idling to wait for passengers, thus reducing the prolonged occupancy of parking spaces in the core area, and thus lowering the demand for on-street temporary parking [58]. Finally, SAVs may form concentrated dispatch parking demand (e.g., standby, charging) in some areas due to idling, etc. If idling is restricted, their parking demand may be concentrated in transportation hubs or core areas [43].

On the supply side, the popularization of SAV may trigger the spatial layout and functional readjustment of urban parking facilities. On the one hand, traditional parking facilities will be reallocated: on-street parking spaces can be converted into bus lanes, bike lanes, sidewalks, or other public spaces, while large parking lots and multi-storey parking buildings located in central business districts have the potential to be redeveloped [59]. On the other hand, the operational characteristics of SAVs will give rise to a new type of parking facility layout pattern, including the installation of low-cost, centralized “hub parking lots” or “dedicated parking zones” on the periphery of the city, and the construction of integrated facilities integrating intelligent dispatch and charging functions to meet the needs of centralized parking of SAVs. The construction of comprehensive facilities integrating intelligent dispatching and charging functions to meet the demand for centralized parking and energy supply of SAVs [60].

Multiple intermediate variables shape SAV’s impact on parking supply-demand, with cross-country comparisons enhancing analytical depth. First, SAV penetration is one of the core variables; high penetration scenarios can significantly reduce overall parking demand, while low penetration has a limited effect on demand reduction [58]. Second, urban density and functional layout affect the spatial distribution of parking, with higher density, mixed-use areas experiencing a greater reduction in parking demand, while lower density, mono-functional areas have a lesser impact [61]. In addition, policy and planning factors have a direct impact on the rate of redistribution of parking supply and land redevelopment [58]. Finally, technology and operation mode determine the spatial distribution of SAVs and their differential impact on parking pressure in core and peripheral areas [60,62].

2.4. Mechanisms of Changes in Supply and Demand of Urban Parking Spaces in the Context of SAVs

Based on the previous research, this study believes that the substitution effect of SAV on private cars may present significant stage characteristics, so it constructs a model of the mechanism of the impact of the parking supply and demand system in three stages: near, middle, and long term. In the initial stage of technology promotion, the prolongation of detention time (The detention time of private vehicles in residential areas refers to the duration from when a vehicle enters the boundary of the residential area and parks at a designated parking space or temporary stopping point, to when it departs from the area.) of private cars may trigger a phased increase in parking demand in the short term. As users’ trust in SAVs increases and their frequency of use increases, their choice of travel mode may gradually shift to shared mobility, thus reducing the actual use of private cars. At the stage of large-scale popularization, driverless cars under the sharing mode have the potential to replace a large number of private cars, which will fundamentally reduce the level of urban car ownership and parking demand (Figure 1).

3. Overview of the Study Area

The parking dilemma of old communities reflects the contradiction of lagging urban planning, and the exploration of spatial renewal can provide a paradigm for the transformation of similar areas. Solving the parking problem is of great significance for improving people’s livelihood, ensuring safety, and optimizing traffic.

Hanyang District, as one of the major urban areas in Wuhan, contains typical unit neighborhoods and old commercial buildings. Parking problems in the district are complicated, especially in the old commercial housing and residential areas face serious parking shortages due to the deficiencies in early construction and planning, while the management is more chaotic, with frequent parking conflicts during peak hours. Meanwhile, Hanyang District, as the first pilot operation area of the “Radish Run” driverless online car, has constructed a driverless dynamic path optimization mechanism based on real-time traffic flow, so that the self-driving vehicles can adaptively integrate into the existing traffic ecology, and has an innovative paradigm of symbiosis between driverless vehicles and urban traffic texture.

Notably, Wuhan and Hanyang have distinct particularities in driverless technology application. Policy-wise, Wuhan is a national intelligent connected vehicle pilot city, issuing policies on road right opening and data security; Hanyang further rolls out targeted support measures for driverless operation, removing barriers for technology landing. In development and popularization, Hanyang has covered core business districts and old residential areas with “Radish Run”, with over 500,000 operating kilometers and serving over 100,000 residents, and its vehicle models are adapted to local traffic scenarios. In terms of acceptance, over 80% of residents recognize driverless safety in local surveys, and communities cooperate with enterprises to hold experience activities, further boosting willingness to use.

This study selected Qintai Community as its case study subject (Figure 2). In China, the definition of old residential communities in historic districts typically refers to residential complexes (including single-building residential units) constructed before the end of 2000 that are neglected, poorly maintained, and lack adequate municipal supporting facilities. High-density communities are generally defined by reference indicators such as a floor area ratio (FAR) of 1.8 or higher, or a population density exceeding 10,000 people per square kilometer.

Built in the early 1990s, this community exemplifies a typical aging open-style neighborhood. According to Wuhan’s 2021 Construction Intensity Management Regulations, Qintai Community in Hanyang District is classified as Zone III with a FAR of 2.3–2.5. Based on existing data, Qintai Community’s population density is approximately 45,600 people per square kilometer. The specific calculation basis is as follows:

Area Data: Qintai Community’s jurisdiction covers 0.15 square kilometers.

Population Data: Interviews indicate a permanent resident population of 6840.

Density Calculation: Population density = 6840 people ÷ 0.15 km² = 45,600 people/km². Both the FAR and population density values significantly exceed standards, reflecting the highly dense population characteristic of old urban communities. Thus, Qintai Community is representative.

Qintai Community exhibits high building density. With 1200 households and approximately 600 private vehicles, aging infrastructure and insufficient parking spaces have led to chaotic parking conditions.

Traffic in the community relies on Hanyang Avenue and Qintai Avenue (with an average daily traffic volume of more than 30,000 vehicles), which are the main roads connecting the Yangtze River Bridge and the Yuehu Bridge, and account for 40% of the transit traffic. During peak hours, traffic on Hanyang Avenue is dense from west to east, and traffic on Qintai Avenue is dense from north to south; on weekdays, the high congestion index reaches 1.8 in the morning and evening, and on holidays, the traffic volume near the Guiyuan Temple, 2 km away, surges, and the congestion extends to the entrance of the community. Residents commute to Zhongjiacun (15%), Wangjiawan (20%), Hankou Wuguang Business District (25%), and Wuchang Zhongnan Road (18%), and non-commuting is concentrated in commercial areas such as Jianghan Road (30%) and Chuhehan Street (22%), forming a tidal wave of traffic characteristics.

The overall width of the road inside the district is 4–6 m, and the mixing of people and vehicles is prominent. There are only 80 parking spaces on the ground, and the average daily parking demand is about 550 vehicles, with the daytime parking space shortage rate exceeding 70%. 18:00 after the street parking space occupancy rate reaches 75%, often squeezing the fire escape. External road congestion interferes with internal microcirculation, with queues at entrances and exits averaging 8 min.

The community has made several attempts to improve the parking status quo: coordinating with Moon Lake Park to provide 150 preferential parking spaces for residents to use temporarily during the construction period; negotiating with neighboring banks to mobilize employees to reduce parking and vacate 42 spaces; hardening of the community’s muddy ground, new parking spaces to improve the parking environment, but the results are very little.

The problem stems from three aspects: First, the mismatch of spatial resources; aging community planning did not anticipate the growth of private cars, parking spaces accounted for a very low percentage of residents crowded fire lanes and green belts and other phenomena are common, and the existence of unused space. Second, the parking management is out of order; the street and residential parking space occupied by the traffic space, in violation of the fire regulations, the external vehicles randomly parked exacerbated congestion. Third, the behavioral inertia of the residents, most of whom are opposed to sharing private parking spaces due to property rights, and the small number of parking spaces at charging stations, which leads to inefficient queuing for charging new energy vehicles.

4. Methodology

4.1. Research Design and Data Source

The design framework of this study consists of four stages (Figure 3). Firstly, in the research background section, from the perspective of urban renewal and community parking dilemma, combined with the theory of shared mobility and the theory of mobility transformation, the questionnaire, model construction, and case study methods were used to clarify the research questions and objectives.

Secondly, in the preliminary research stage, through on-site observation, enterprise interviews and public perception surveys, combined with the pilot experience of “Robo taxi”, the Qintai community in Hanyang District of Wuhan City is selected as a case study to carry out a systematic survey on community distribution, parking demand, traffic environment, and parking supply.

Thirdly, in the data analysis stage, the comprehensive questionnaire data analysis and interview data are used to measure the trend of community parking demand, focusing on the analysis of residents’ willingness to purchase cars, parking space turnover rate and parking behavior patterns, and then predict the evolution of parking demand in the near, medium and long term. Since the promotion of driverless technology has stage characteristics, policy planning is usually promoted in a five-year cycle, and residents’ acceptance of the new technology shows a gradual process from cognition, trial to popularization, this study sets three stages in the questionnaire design, namely, the next 5 years, 5–10 years, and more than 10 years, to predict the evolution of private car ownership under the popularization of SAV in a more scientific manner [19].

Finally, in the strategy proposal stage, based on the empirical results and trend prediction, differentiated phased parking optimization strategies are proposed, including short-term contingency measures, medium-term facility renovation solutions, and long-term governance and operation mechanisms, to provide theoretical and practical references for the balance of parking supply and demand and the efficient use of resources in the community.

The data collected in this study include interview information and questionnaire data. The data on residents’ car ownership, current parking supply, vehicle detention rate at different times of the day, average household visit rate, and the amount of borrowed parking by neighboring enterprises and institutions come from the interview records with community staff (Appendix B). Data on residents’ SAV awareness levels, travel intentions, travel characteristics, and rates of change in vehicle ownership were derived from survey results (Appendix A).

In conducting the community survey on residents’ SAV acceptance, 148 questionnaires were distributed, and 135 were validly returned. In this study, the sample group was mainly targeted at residents in the age group of 26–40 years old, who are mainly employees of enterprises and institutions and freelancers, and whose household SAV ownership is mainly 1 vehicle. This design is based on three considerations: first, the 26–40 age group is in the peak period of professional and family responsibilities, and is the core group with the most diversified and active urban travel demand. In Wuhan, drivers aged 26–50 account for 71.8% of the total, and shifts in their travel behavior are particularly critical for the promotion of SAVs, the impact of their travel mode change on the promotion of SAV is the most critical [16,63]. Secondly, employees of enterprises and public institutions and freelancers generally have stable economic resources and strong consumption decision-making ability, which can better reflect the potential willingness to pay for SAVs and the substitution effect of vehicle purchase [57]. Finally, Chinese urban households generally show the ownership structure of “one vehicle per household” [41], and this group faces the choice between “continuing to purchase more vehicles” and “relying on shared mobility” in the future, and their acceptance of SAVs is important for replacing private vehicles. This group is facing the choice between “continue to purchase more” and “rely on shared mobility” in the future, and their acceptance is typical of the demand for SAVs to replace private cars. Therefore, the above sample selection can ensure that the survey results can fully reveal the conceptual change and behavioral intention of the major travel groups in the context of driverless shared mobility, and provide a reference for subsequent policies and planning.

4.2. Survey on Residents’ Acceptance of SAVs

To systematically assess residents’ perceptions, attitudes, and potential behavioral responses to shared driverless vehicles, this paper constructs a survey framework for residents’ acceptance of SAVs based on established research results (see

Table 1. Indicator table of residents’ acceptance of the SAV survey.

First-Level Dimension	Second-Level Dimension	Specific Indicators	Description
Basic Information	Gender	--	--
	Age	--	--
	Occupation	--	--
	Residential environment	Area type	Urban (≤5 km)/Suburban (>5 km)
		Neighborhood type	Gated community (unit housing/commercial housing)/open neighborhood (open community)/other (open text)
Status of Travel Behavior (work–life categorization discussion)	Transportation Resources	Private car ownership	Yes (household ≥ 1 car)/No
		Parking space status	Fixed owned parking space/leased parking space (monthly payment)/no fixed parking space
	Commonly Used Mode of Transportation	Main mode of commuting	Private car/Internet car/bus/subway/others (according to the most frequent judgment)
		Main Mode of Travel for Daily Life	Private car/Internet car/bus/subway/others (according to the most frequent judgment)
	Travel Characteristics	Main commuting distance	≤3 km/3–5 km/>5 km
		Bus stop accessibility	≤10 min walk/10–30 min walk/>30 min walk
	Parking Problems	Parking Difficulty Perception	Insufficient parking spaces/excessive parking fees/parking indiscriminately/exotic cars occupying spaces/lack of charging piles/others
		Adequacy of Parking Spaces	1 (Insufficient)–5 (Sufficient)
Attitude of SAV Cognition	Technology Familiarity	SAV Awareness	Know very well/know some/have heard of it/know nothing at all
		Driverless experience	Yes/No
	Psychological acceptance	Level of trust in the technology	Likert scale of 5: 1 (very distrustful)—5 (very trusting)
		Safety Concerns	Likert 5 score: 1 (very worried)—5 (very reassuring)
		Willingness to try it for a short period of time	Very willing/More willing/Depends/Less willing/Very unwilling
SAV travel impacts	Willingness to substitute	Future Travel Preferences	SAV dominated/mixed mode/private car dominated
		Willingness to buy a car impact	Significant decrease/likely decrease/no impact/likely increase
	Cost sensitivity	Single-trip cost premium acceptance	≤current cost/+10–20%/+20–30%/>30%
		Time and Cost Preferences	Prioritize time savings/prioritize money savings/balance.
Medium to Long Term Expectations	Utilization Planning	Likelihood of use within 5 years	Likert scale of 5: 1 (not at all likely)—5 (very likely)
		Willingness to give up private car in 10 years	Yes/No/Unsure
	Expected service scenarios	Expected Service Scenarios	Commuting/night travel/transportation needs/shopping and entertainment/cross-district travel/traveling after drinking alcohol/others
	Expectation of Parking Improvement	Parking pressure improvement expectation	Significant Improvement/Partial Improvement/No Impact/Deterioration

). The framework mainly consists of five first-level dimensions: basic information, current status of travel behavior, SAV cognitive attitudes, SAV travel impacts, and medium- and long-term expectations, which are measured with 24 second-level indicators. The items of the five scales were self-developed based on relevant literature and preliminary interviews with residents, to ensure contextual relevance to the study area.

Among them, the basic information and travel characteristics sections are used to portray the basic attributes and daily travel characteristics of the sample, providing control variables for subsequent group comparisons. The status quo of travel behavior includes indicators of private car ownership, parking conditions and travel purpose, and travel mode choice, which are used to measure the structure of residents’ transportation and travel dependence [25,43]. The SAV cognitive attitudes section is designed to measure Likert scale measures of safety, comfort, psychological security, and technological trust, which are designed to capture residents’ psychological acceptance of the emerging travel technologies [62,64,65]. The SAV travel influences section focuses on The SAV travel impact section focuses on willingness to purchase a car, travel preferences, and price sensitivity, reflecting the possible behavioral shifts of residents after the popularization of SAV [65]. The medium and long-term expectations assess the potential impact of SAVs on parking demand and the transportation system by projecting travel patterns and private car ownership 5–10 years in the future and 10 years from now [19].

Through this framework, a systematic measurement of residents’ cognitive-attitudinal-behavioral intentions can be achieved, providing an empirical basis for subsequent quantitative analysis of the role of SAV in alleviating parking pressure, influencing private car ownership, and optimizing urban transportation.

4.3. Community Parking Demand Measurement Methodology

4.3.1. Core Formula for the Number of Community Parking Spaces

It is extremely important to differentiate between periods in the community parking demand assessment. First of all, there are significant differences in the ratio of residents’ vehicles to visitors’ demand during different periods: residents may travel out during working hours, and visitors’ demand increases; while residents’ vehicles return intensively at night, and visitors’ demand decreases relatively; during holidays, there may be a high demand for visitors superimposed on residents’ long-term parking situation. Second, a large number of empirical studies have emphasized that urban parking demand has a clear time-space pattern. For example, parking use in different administrative lots is characterized by diverse “rhythmic” characteristics and is significantly correlated with the degree of land use mix in the vicinity [66]. Therefore, in order to realize the accurate allocation and management of parking resources, it is necessary to adopt differentiated measurement and planning strategies for different periods. This study distinguishes the following three periods: weekday working hours (e.g., 8:00–18:00), weekday resting hours (e.g., 18:00–8:00 the next day), and holiday/weekend hours. The demand for parking is calculated for each of the three periods and compared to the supply: if the supply exceeds the current level, it is considered that there is a parking shortage, and if it is less than or equal to the current level, it is considered that there is no parking shortage. Through the comparison to explore the changes in parking supply and demand. The specific calculation process is shown in Figure 4.

In community parking demand assessment, theoretical and empirical studies have consistently shown that the total parking demand should be integrated with the needs of multiple user groups, including residents’ long-term parking, visitors’ short-term parking demand, and redundant parking spaces (often referred to as “spare” spaces) reserved to cope with fluctuations in demand [67,68]. In summary, the calculation of total community parking demand for all three time periods in this study follows the formula [69]:

Total Community Parking Demand =
Resident Parking Demand + Visitor Parking Demand + Redundant Parking Spaces

(1)

4.3.2. Itemized Calculation Formulas

• Resident Parking Spaces:

Relevant studies suggest that residential parking demand is determined not only by vehicle ownership within a community but also by variations in vehicle retention rates across different time periods [70]. Higher car ownership generally corresponds to greater parking demand, while fluctuations in retention rates during peak hours and overnight parking directly shape the dynamics of parking space shortages. The calculation formula is presented as follows [69].

Resident parking space demand = (resident car ownership × rate of change
of car ownership) × different time period detention rate

(2)

where the rate of change of car ownership is estimated based on the indicators of the two dimensions of SAV travel impact and medium- and long-term forecast in the questionnaire survey.

2.

Visitor Parking Spaces

According to Kumakoshi, Y. et al., the demand for visitor parking spaces is not only directly related to the number of visitor vehicles but is also influenced by the turnover rate of parking spaces, which varies across different time periods [62]. For instance, during working hours on weekdays, visitor vehicles often include temporary parking from nearby enterprises and institutions and are characterized by short-term stays, whereas during non-working periods, visitor vehicles are more likely to be parked for long-term. These differences contribute to the temporal variability of visitor parking demand. The calculation formula is as follows.

Visitor parking space demand =
average daily number of visitors vehicles/turnover rate

(3)

Number of visitor vehicles during weekday work hours = number of households ×
average household arrival rate + borrowed parking from neighboring businesses and institutions

(4)

Number of visitor vehicles during weekday rest time = number of households ×
average household visit rate

(5)

Number of visitor vehicles on rest days = number of households × average household visit rate

(6)

“Turnover rate” represents the number of times a single parking space is used by different vehicles in a unit of time (day), and the turnover rate is determined by the average length of stay of visitors (according to the field research, this study adopts a stay of 2 h and is open 10 h during weekday work hours, 14 h during weekday rest time and all day long during weekend time, so the turnover rate is = 24/2 = 12 times/day) [71].

3.

Redundant parking spaces

Redundant parking spaces =(resident’s parking spaces + visitor’s parking spaces + ancillary parking spaces) × β

(7)

Redundant parking spaces are usually set up as a “back-up” to actual demand, and are designed to handle the overlapping demand of residents, visitors, and amenities during peak hours [57]. In this study, the redundant parking spaces are the parking spaces for non-residents and non-visitors, such as commercial, educational, and medical services. etc. β is the redundancy coefficient, which is usually 10% to 20%. In this study, the target population is a community of flats, where parking spaces are in short supply, and the redundancy factor is calculated at 10% [67].

5. Results

Cronbach’s α reliability tests were conducted for the three dimensions of the residents’ SAV acceptance survey scale—travel characteristics, residential characteristics, and trust and acceptance—and the results are shown in

Table 2. Reliability and validity testing of questionnaire data.

Dimension	Cronbach’s α	KMO Value	Approx. χ2	p
Travel characteristics	0.885	0.771	2472.828	0.000
Residential characteristics	0.600	0.654	71.793	0.000
Trust and Acceptance	0.736	0.734	166.993	0.000
Overall	0.914	0.800	3318.368	0.000

. As indicated in the table, Cronbach’s α values for the three dimensions were 0.885, 0.600, and 0.736, respectively, all greater than or equal to 0.6. Since the α values did not increase after deleting any individual item, the internal consistency within each dimension is considered acceptable, indicating satisfactory reliability.

To verify the structural validity of the questionnaire, the KMO test and Bartlett’s test of sphericity were conducted, and the results are presented in Table 2. As shown, the KMO values for the travel characteristics dimension and the trust and acceptance dimension were 0.771 and 0.734, respectively, both greater than 0.7, suggesting suitability for exploratory factor analysis. The KMO value for the residential characteristics dimension was 0.654, within the acceptable range of 0.6–0.7; considering the limited number of items, it still meets the basic requirements for factor analysis. Bartlett’s test of sphericity showed approximate chi-square values of 2472.828, 71.793, 166.993, and 3318.368 for each dimension and the overall scale, with all significance levels at p = 0.000 (p < 0.001), indicating significant correlations among the items.

5.1. SAV Awareness Levels of Residents in Qintai Community

5.1.1. Acceptance of Unmanned Technology by People of Different Occupational Types

In the early stage of SAV popularization, the public maintained a high level of concern and interest. The survey results show that the public as a whole holds a high acceptance of driverless technology, but the differences between different occupational groups are more significant. As shown in Figure 5, among those who do not accept driverless technology, the proportion of freelancers is higher; among those with average acceptance, the proportion of each occupational group is relatively balanced; and among those with high acceptance, the proportion of employees of enterprises and public institutions is significantly higher than that of other groups. And the majority of the interviewed residents are from employees of enterprises and public institutions, indicating a high acceptance of autonomous driving in the community.

5.1.2. Trust in Unmanned Technology Among People of Different Occupational Types

In the early stage of SAV’s popularization, the public’s trust is high. The survey results show that the public’s trust in driverless technology as a whole is dominated by “general” and “trust”, and the number of both is basically the same, with only a few respondents saying “no trust”. Only a few respondents said they “don’t trust”. Further analysis by occupational grouping reveals that there are some differences among different occupational groups. In the group of “distrust”, the proportion of freelancers and retired and other people is higher, followed by employees of enterprises and public institutions, while the proportion of students is lower; in the group of “general trust”, the distribution of various occupational types is relatively balanced; in the group of “trust”, the number of respondents is the same, with only a small number of respondents saying “trust”. In the group of “trust”, employees of enterprises and public institutions accounted for the highest proportion, followed by students. Overall, people who may rely on cars for their daily trips have an “average” or “trust” attitude towards driverless technology, with only a small number of respondents showing distrust due to uncertainty about the new technology. This result suggests that, from the perspective of trust, driverless technology as an alternative to private vehicles is feasible among the public (Figure 6).

5.1.3. Travel Choice Preferences of Different Types of People

At the early stage of SAV popularization, residents’ preference for driverless travel is mainly focused on high-frequency commuting and feeder, and scenarios such as high-risk, long-distance and special conditions, while acceptance in low-frequency or complex road conditions is relatively low. The survey results show that there are differences in the choice of driverless application scenarios among different occupational groups. In high-frequency scenarios such as “daily commuting” and “public transportation connection”, students and employees of enterprises and public institutions have a higher proportion of choices; in “long-distance travel” and “night travel”, the proportion of students and employees of enterprises and public institutions is higher. Students and employees of enterprises and public institutions have a higher proportion of choice in high-frequency scenarios such as “long-distance travel”, “nighttime travel” and “remote suburbs”; in scenarios with stronger safety and convenience needs such as “rainy and snowy weather” and In special environments such as “rainy and snowy weather” and “complex road conditions”, the proportion of employees of enterprises and public institutions choosing is relatively higher; in substitute scenarios for chauffeur-driven vehicles such as “traveling after drinking alcohol” and “picking up and dropping off at airports/stations”, the acceptance rates of students and employees of enterprises and public institutions are relatively higher. In the alternative scenarios of “traveling after drinking” and “airport/station pickup”, the acceptance of students and employees of enterprises and public institutions is outstanding. Overall, the public is more inclined to choose driverless vehicles in high-risk, long-distance, and special conditions, while the acceptance of driverless vehicles in low-frequency or complex road conditions is relatively low (Appendix C).

5.1.4. Summary of Results

Overall, the public has high acceptance of driverless technology. Employees of enterprises and public institutions account for the largest share of high acceptance—and most interviewed residents belong to the latter group, reflecting high acceptance of autonomous driving in communities. In terms of trust, employees of enterprises and public institutions and students who will be the main driving population in the next five years have higher trust, indicating that driverless technology is feasible as an alternative to private vehicles. For travel choice preferences, residents prefer driverless travel in high-frequency scenarios.

To sum up, SAV has a good public foundation for promotion, especially among employees of enterprises and public institutions and students. It is feasible to promote it as an alternative to private vehicles, and future promotion can focus on high-frequency and special travel scenarios while paying attention to guiding groups with relatively conservative attitudes to enhance their recognition of unmanned technology.

5.2. Travel Intention and Characteristics of Residents in Qintai Community

The survey shows that residents in the Qintai community mainly use public transportation for commuting, and rely more on online rides for daily flexible trips, while the use of private cars is relatively low due to parking constraints. Residents of the Qintai community show significant differences in different travel scenarios. For commuting to work and school, the use of public transportation is the highest (70.4% 3–5 times per week), with one-way time mainly concentrated in less than 15 min and 30–60 min, and the cost of commuting mainly in the range of RMB 50–100; the use of private cars is relatively low (44.4% do not use them), and the use of internet rides is extremely rare (90.4% do not use them). As for daily life trips (shopping, recreation, leisure, etc.), the use rates of public transportation and private cars are low, and Net Journey has become the main mode of travel (66.7% use it, 53.3% use it 3 times or more per week), with one-way trips mostly lasting 15–30 min and weekly costs mostly in the range of RMB 50–100, showing that residents prefer instantaneous services in flexible travel. In terms of residential location characteristics, the neighborhoods are open communities with predominantly older commercial properties, closer to downtown and major commuting destinations, and convenient walking distance to transit stops, but parking facilities are tight. Overall, residents’ commute trips emphasize cost control, while their daily life trips rely more on flexible and instantaneous transportation modes, which are also limited by the living environment and parking conditions.

In terms of parking demand, according to Table 2, the existing parking spaces are about 200, and the parking demand of residents is about 238, which is a problem of insufficient parking spaces (

Table 3. The value of each index is used for the calculation of the residents’ parking demand in stages.

Item	Formula	Assumptions
Resident Parking Demand	Demand = Resident Car Ownership × Rate of change of car ownership × Time-based Occupancy Rate	Current resident Car Ownership. 400	Occupancy rates. Weekday work hours: 30%. Weekday rest hours: 60%. weekend: 50%.
		Rate of change of car ownership: First stage: around 1 Next 5–10 years: 0.9 10 years later: 0.7
Visitor Parking Demand	Demand = Daily Visitors: Turnover Rate	Calculation basis. Weekday work hours: 0.005/household Weekday rest hours: 0.01/household Weekend: 0.02/household
Visitor Parking Turnover	Turnover Rate = Open duration ÷ Average Parking Duration	Assumed parking durations: Weekday work hours: 5 cycles/day Weekday rest hours: 7 cycles/day Weekend: 12 cycles/day
Redundancy Parking	Redundancy = (Resident Parking + Visitor Parking) × β	β = 1.1 (Redundancy factor: 110–120%)

).

Residents’ preference for flexibility and immediacy in daily travel is highly aligned with the characteristics of shared autonomous driving—on-demand deployment and no need for fixed parking spaces. Meanwhile, shared autonomous driving adopts a “multi-user shared-vehicle” model, which eliminates the need to allocate fixed parking spaces for each user. Relying on existing road resources, it supplements the demand for flexible travel, further reduces residents’ reliance on private cars, and alleviates parking pressure at the source.

5.3. Stage-by-Stage Forecast of Changes in Parking Demand for Residents in the Qintai Neighborhood

5.3.1. Stage 1—Parking Space Becomes More Scarce

1.

Travel mode choice

The survey shows that more than half of the residents have an open or partially accepting attitude towards driverless car sharing in the roughly 5 years to come. And that there are differences in the acceptance of driverless car sharing by residents: in the group without private cars, 8.9% are willing to rely on it completely, and 38.5% are willing to use it partially; in the group with private cars, 10.4% are willing to replace it completely, and 20.0% partially use it; and 22.2% do not consider it at all. Overall, most of the residents have an open or partially accepting attitude towards driverless car sharing, showing its future travel potential (Figure 7).

2.

Calculation: Total Parking Demand in Qintai Community in Stage 1

During the first stage, the contradiction between residents’ parking supply and demand has intensified. According to

Table 4. Calculation of current parking demand in the Qintai neighborhood.

Item	Calculation Process
Weekday work hours parking demand	143 spaces
Weekday off-hours parking demand	315 spaces
Weekend parking demand	228 spaces
Average parking demand	238 spaces

, the residents’ parking demand in the first stage is about 275 spaces, the current parking demand is about 238, which is an increase of about 15.55% year-on-year in demand. And the parking supply is around 200 spaces, which may not change much in the future (Appendix B). It makes the parking space even tighter. From the questionnaire data (pre-survey selection tendency and the above results), it is concluded that due to the increasing trend of unmanned by replacing some of the trips, the residents’ car ownership may decline negligibly in the first stage, but the rate of stranding on weekdays during the working hours may increase to 50%, the rate of stranding on weekdays during the rest hours may increase to 70%, and the rate of stagnation on days off may increase to 60%. Visiting vehicles may decline by about 30%.

5.3.2. Stage 2—Parking Space Constraints Are Alleviated

1.

Travel mode choice

The survey results show that more than half of the residents have an open or partially accepting attitude towards driverless car sharing in the next approximately 5–10 years. The acceptance of driverless car sharing among residents shows a clear division: among the group without private cars, 12.6% are willing to fully rely on driverless car sharing, and 39.3% are willing to partially use it; among the group with private cars, 12.6% are willing to completely replace private cars and 23.0% are willing to partially use them; another 12.6% of residents do not consider driverless shared cars at all. Overall, more than half of the residents have an open or partially accepting attitude toward driverless shared cars, indicating that they have greater potential in medium- and long-term travel planning (Figure 8).

2.

Calculation: Total Parking Demand in Qintai Community in Stage 2

In the next 5–10 years, the contradiction between residents’ parking supply and demand will be alleviated. According to

Table 5. Calculation of Parking Demand in the Neighborhood in Stage 1.

Item	Calculation Process
Weekday work hours parking demand	228 spaces
Weekday off-hours parking demand	313 spaces
Weekend parking demand	269 spaces
Average parking demand	275 spaces

, the parking demand of residents in the next 5–10 years is about 208 cars, the current parking demand is about 238, and the current parking supply is around 200 vehicles, which is about 12.6% lower than the current year demand but basically equal to the supply of parking spaces, which may not change much in the future (Appendix B). The shortage of parking space has been alleviated. From the questionnaire data (pre-survey selection tendency and the above results), it is concluded that due to the increasing trend of unmanned by replacing some of the trips, the residents’ car ownership in the next 5–10 years may decline by 10–20% compared to the status quo, taking 10%, and the weekday work period stall rate may rise to 40% compared to the status quo, the weekday rest period stall rate may rise to 60%, and the rest day stagnation rate may rise to 50%. Visiting vehicles are likely to remain the same over the first stage.

5.3.3. Stage 3—Limited Parking Space Fades Away

1.

Travel mode choice

The survey results show that in 10 years, the vast majority of residents have an open or accepting attitude towards driverless car sharing. Among the group without private cars, 34.1% are willing to fully rely on driverless shared cars and 23.0% are willing to partially use them; among the group with private cars, 25.9% are willing to fully replace their private cars with driverless shared cars and 14.8% are willing to partially use them; and only 2.2% of the residents do not consider driverless shared cars at all. Overall, the vast majority of residents have an open or accepting attitude toward driverless shared cars, showing that they have broad application potential in the future long-term travel system (Figure 9).

2.

Calculation: Total Parking Demand in Qintai Neighborhood in Stage 3

After 10 years, the problem of limited residential parking spaces will gradually disappear as the parking supply-demand contradiction eases. According to

Table 6. Calculation of Parking Demand in the Neighborhood in Stage 2.

Item	Calculation Process
Weekday work hours parking demand	166 spaces
Weekday off-hours parking demand	242 spaces
Weekend parking demand	203 spaces
Average parking demand	208 spaces

and

Table 7. Calculation of Parking Demand in the Neighborhood in Stage 3.

Item	Calculation Process
Weekday work hours parking demand	68 spaces
Weekday off-hours parking demand	127 spaces
Weekend parking demand	96 spaces
Average parking demand	100 spaces

, the parking demand of residents after 10 years is about 100 spaces, and the current parking supply is around 200 spaces that will not change much in the future (Appendix B), which is about 58.0% lower than the same period of the previous year’s demand (208) and the problem of limited parking space is gradually disappearing. From the questionnaire data (pre-survey selection tendency and the above results), it is concluded that due to the increasing trend of unmanned by replacing some of the trips, the residents’ car ownership may decline by more than 30% compared to the status quo after 10 years in the future, take 30%, and the rate of stalling on weekdays during the working hours may decline to 20% compared to the status quo, the rate of stalling on weekdays during the rest hours may decline to 40%, and the rate of stalling on rest days may decline to 30%. Visiting vehicles may decrease to 50% from the status quo.

5.3.4. Summary

Overall, parking pressure rises in the short term, stabilizes in the medium term, and declines significantly in the long term. Based on this, the program design can be divided into three phases: short-term (next 5 years) to respond to peak growth and achieve staggered parking by sharing parking resources with shopping malls, office buildings, etc.; medium-term (5–10 years) to transition and optimize, enhance public transportation and bicycle facility connections, and reduce reliance on private vehicle parking; and long-term (more than 10 years) to proactively reduce and transition parking and promote a walk-first car-free community design that prioritizes walking and reduces parking demand overall.

6. Discussion

6.1. Factors Influencing Resident SAV Acceptance

Existing studies indicate that SAVs, owing to their advantages of high safety, convenience, and low cost, may exert a significant substitution effect on private cars (e.g., one SAV can potentially replace 10–13 private vehicles), thereby continuously reducing overall community parking demand [24,25,26]. Meanwhile, several studies have indicated that in the absence of effective regulation and pricing mechanisms, autonomous vehicles may substantially increase travel demand by reducing travel costs, enabling additional trips for non-drivers, and generating empty vehicle mileage. These effects could consequently lead to an overall rise in vehicle miles traveled (VMT) and exacerbate traffic congestion [72,73]. It has been suggested that, without corresponding policy interventions, the widespread adoption of privately owned autonomous vehicles could result in a 4–20% increase in VMT, underscoring the necessity of implementing measures such as mileage-based charges and congestion pricing to mitigate these impacts [74]. Nevertheless, the actual evolution of residents’ parking demand may follow a nonlinear trajectory. Specifically, with policy incentives, technological advances, and the wider adoption of SAVs, residents in Qintai Community are expected to gradually enhance their acceptance and trust in this new mobility mode. In the short term, some residents may partially substitute their travel with SAVs, leading to increased idle time of existing private cars, which in turn exacerbates the shortage of community parking spaces. As the technology becomes more prevalent, a growing number of residents are likely to fully replace private cars with SAVs, gradually alleviating the problem of parking scarcity. In the longer term, community parking spaces will be further released, and the newly available land can be repurposed for public activities, thereby contributing to easing the urban issue of parking difficulties.

The survey results show that the main problems of the current situation in the community include insufficient parking spaces, high parking fees, occupancy by foreign vehicles and lack of charging piles, indicating that the conflict between supply and demand and insufficient management together affect the residents’ travel experience, and that the residents’ behavior of choosing SAVs instead of private cars is affected by a variety of mechanisms.

The survey results indicate that residents’ choice of SAVs is mainly influenced by four factors: safety and trust, as they seek to avoid risks associated with human driving [75,76]; perceived efficiency, as they prefer the dynamic scheduling of SAVs over the high idle rate of private vehicles [77]; spatial and cost benefits, as they perceive the release of parking space and reduction in commuting costs; and the policy environment, as government pilot programs and publicity efforts enhance public acceptance [78]. Overall, residents employed in enterprises or public institutions show higher acceptance levels, while students and freelancers exhibit greater resistance, reflecting the influence of occupational differences on technological trust.

6.2. Suggestions for Parking Space Optimization in the Context of Urban Renewal

6.2.1. Phase I: Space Optimization and Sharing Mechanism

In the initial stage, when parking spaces become increasingly scarce, the overall vehicle retention rate of community residents rises temporarily. Accordingly, strategies during this period should focus on alleviating short-term parking shortages. Taking Qintai Community as an example, the first measure is to promote “staggered shared parking.” By leveraging the community’s locational advantages, parking resources in surrounding commercial complexes, office areas, and residential neighborhoods can be coordinated for time-based sharing—for instance, allowing residents to use commercial parking spaces at night [79]. In addition, vacant buildings in Qintai Community may be transformed into intelligent multi-level parking facilities [80]. At the same time, curbside parking can be subject to intelligent management through AI-enabled dynamic pricing systems. Collectively, such measures enable more efficient allocation and utilization of parking resources without significantly increasing land supply [81].

In this process, residents’ opinions and participation are crucial. Through a bottom-up governance mechanism, such as a residents’ deliberation committee—a classical approach in Chinese community governance—community members can reach consensus on issues including parking space allocation, pricing policies, and first-come-first-served rules. They can also provide feedback during the design, implementation, and monitoring stages of management plans. This approach ensures that parking management measures align with the actual needs of the community, enhances residents’ sense of involvement and satisfaction, and offers practical insights and evidence for urban renewal and the governance of public spaces.

6.2.2. Stage 2: Structural Adjustment in Transition

During the transition period, community parking faces several challenges. Although private car ownership declines, it is not fully replaced by SAVs, leading to an incomplete reduction in demand and a spatial mismatch between redundant and scarce parking spaces. The coexistence of SAVs and private cars also creates pressure on existing road networks, raising congestion risks and reducing the efficiency of parking use. Moreover, the mix of travel modes makes parking demand unstable, adding uncertainty to planning and management. Therefore, addressing these challenges calls for coordinated efforts in both infrastructure transformation and business model innovation to ensure efficient use and orderly transition of community parking spaces.

Carry out multifunctional transformation of surface parking lots, converting 30% of excess parking spaces into charging stations, green logistics distribution stations or recreational community spaces; deploy V2X systems to build a digital road network, and conduct precise route planning via information interactions to serve the driving of SAVs and efficient use of parking spaces [82]; promote MaaS platforms, integrate public transportation, shared self-driving vehicles and micro-circulation buses to advance efficient planning of parking spaces [83], and develop securitization of underground space properties, allowing owners to convert unused parking spaces into investment income through real estate investment trust funds and attract social capital to participate in the construction and operation of parking spaces [84].

6.2.3. Stage 3: Transformation of Urban Form and Community Functions

In the third stage of the limited parking space problem, the demand for private cars continues to decline, and the program will begin to focus on the deeper transformation of urban form and community functions. In terms of the transformation of urban form, the first is to develop flexible parking standards, requiring new buildings to be built as “dynamic mobility service centers” and equipped with self-driving vehicle hubs, abandoning traditional fixed parking standards in favor of on-demand parking facilities. Secondly, the road resources are reconfigured by changing a four-lane road into a two-lane road with dedicated self-driving lanes and multi-functional shoulders, where the dedicated lanes guarantee the efficient driving of SAVs, and the multi-functional shoulders can be used for parking when necessary, thus realizing the multi-functional utilization of the road resources [80]. In terms of community function regeneration, on the one hand, the parking lot is transformed into a community service complex; on the other hand, the Mobile Carbon Account is applied to record the residents’ green travel data based on the blockchain, with which residents can redeem the benefits of the community services, encouraging green travel, and optimizing the parking space from the demand side [85].

7. Conclusions

This paper centers on the core proposition of “whether shared autonomous vehicles (SAVs) can alleviate the urban parking problem”, and develops the research with the logical main line of “Problem identification—Theory construction—Empirical analysis—Strategy proposal “as the logical main line to carry out the research. Taking the contradiction between urban parking supply and demand as an entry point, this study focuses on the potential impact of SAV technology on parking space, aims to predict the change of private car ownership and the degree of SAV substitution in different periods, examines the role of SAVs in parking optimization, and puts forward a staged solution. The main results of this study are as follows:

1.

Mode of travel and parking demand: residents of the Qintai community mainly use public transportation for commuting, and the use rate of private cars is low, mainly due to parking constraints; for daily flexible travel, they rely more on online car rental, and prefer instant and flexible services in daily life travel. The average parking demand of residents is about 238 parking spaces, while the existing parking spaces are only about 200; there is an obvious gap, reflecting the outstanding contradiction between parking supply and demand in the community.

2.

Residents’ awareness and acceptance of SAVs: The public as a whole has a high level of acceptance of driverless technology, but there are differences among occupational groups. Students and freelancers are more resistant to the new technology, while employees of enterprises and public institutions account for a larger proportion of the high acceptance group, indicating that they are more positive about the potential application of SAVs.

3.

Residents’ trust in unmanned vehicles and their preferences for usage scenarios: “general” and “trust” are dominant, with only a few showing distrust. The public prefers SAVs in high-risk or high-convenience scenarios such as high-frequency commuting, long-distance travel, nighttime travel, and substitute for chauffeur-driven vehicles, and accepts them less in low-frequency or complex road condition scenarios.

4.

Empirical evidence shows that community parking pressure follows a “short-term increase (initial stage of SAVs, demand rises by 15.5%)—medium-term stabilization (transition stage, demand decreases by 12.6%)—long-term decrease (widespread adoption of SAVs, demand drops by 58%)” trend. Corresponding optimization strategies should be implemented in phases: in the short term, focus on efficiency through “sharing + intelligence”; in the medium term, emphasize adjustments in “facilities + mobility modes”; and in the long term, achieve transformation in “space + function.” Compared with the high idling rate and waste of resources of private cars, SAVs can significantly reduce lanes and parking demand under the sharing mode, lower commuting and car maintenance costs, and realize the substitution of one car for multiple private cars to meet the demand for low-cost, safe, and efficient travel.

5.

Parking optimization in the context of urban renewal needs to be promoted in stages: The stage 1 focuses on easing tension and improving efficiency through sharing, smart garages and dynamic management; the stage 2 combines the popularization of SAVs, promotes the transformation of parking spaces and smart road networks to optimize the structure; and in the stage 3, with the decline in demand for private cars, implements flexible parking standards, reconfigures the functions of roads and communities, and realizes the transformation of parking space to comprehensive services and green travel incentives. Stage 3 is the implementation of flexible parking standards with the decline of private car demand, and the reconstruction of road and community functions to realize the transformation of parking space to comprehensive services and green travel incentives.

The core innovations of this study are reflected in two aspects. Firstly, against the background of urban regeneration, it breaks away from the traditional mindset of “simply increasing parking spaces” to address parking issues in aging communities—a approach that often faces constraints due to limited space and outdated infrastructure. For the first time, the study dynamically links residents’ acceptance of Shared Autonomous Vehicles (SAVs) with community parking demand, and proposes a new perspective of “alleviating parking dilemmas through the innovation of travel modes.” This provides a solution path for aging communities like Qintai Community (characterized by aging facilities and space limitations) that goes beyond physical space renovation, effectively bypassing the bottleneck of traditional parking improvement methods. Secondly, in line with the development trend of artificial intelligence (AI), the study leverages big data technology to integrate multi-dimensional data, including residents’ travel trajectories and parking time preferences. By combining AI-based prediction models to optimize the deployment scale of SAVs and the allocation plan of parking resources, it achieves accurate matching between demand prediction and resource allocation, thereby significantly enhancing the feasibility and efficiency of the proposed solution and addressing the long-standing issue of mismatched supply and demand in parking resources.

In terms of forward-looking significance, this study provides a paradigm for the integration of “smart transportation + community governance” in future urban regeneration. As AI technology continues to advance—such as the integration of real-time traffic condition coupling and multi-modal travel coordination—the application scenarios of this research can be further expanded, laying a foundation for the integrated development of smart city infrastructure. Its key contributions are as follows: theoretically, it fills the research gap in the cross-disciplinary study of SAVs and community parking demand within the context of urban regeneration; practically, it offers actionable solutions that balance residents’ acceptance and parking efficiency for aging communities in Wuhan and other similar cities, providing strong support for the coordinated development of community regeneration and smart transportation, and offering valuable references for policy-makers and urban planners in similar contexts.

Although this study uses Wuhan as a case study, its findings hold broader applicability. As the core platform for SAVs deployment, Baidu’s “Robotaxi” is no longer exclusive to Wuhan. By July 2024, it had launched fully driverless autonomous operations across 12 cities including Beijing, Shanghai, Guangzhou, Shenzhen, and Chongqing, achieving full coverage of China’s top-tier metropolises. From a developmental perspective, SAVs are accelerating their penetration into more major cities. At the policy level, cities like Beijing and Shanghai have introduced specialized regulations to expand testing and operational scopes. The industry anticipates that related services will extend to 65 cities by 2025. Wuhan’s operational experience as the world’s largest autonomous driving service zone—including its residents’ acceptance patterns and evolving parking demand dynamics—can provide valuable reference points for other major cities advancing SAVs pilot programs. This offers practical insights for anticipating traffic space optimization directions in similar urban renewal projects.

However, this study still has limitations. Study takes the Wuhan Qintai community as a single case, whose aging, high building density, and lack of parking facilities are typical, but it is difficult to cover other types of samples, such as newly built gated communities and suburban low-density communities. Due to the heterogeneity of different cities in terms of transportation policies, residents’ travel behaviors, and the progress of SAV technology implementation, further validation is needed for the cross-regional and cross-type dissemination of the research findings. Furthermore, the study focuses on China’s domestic context and the specific scenario of Wuhan’s Qintai Community, neglecting the unique characteristics of different countries in areas such as transportation systems like reliance on private vehicles and public transit coverage in some developed nations, resident travel culture like varying trust thresholds for autonomous driving technology), and SAV industry policies like government subsidy levels and technical standards systems. Consequently, its conclusions hold limited reference value in international contexts addressing topics like “urban parking optimization” and “acceptance of autonomous mobility.” Consequently, these findings hold limited reference value in international contexts and cannot be directly applied to community studies in other countries.

It should be noted that, due to the age structure of residents in the Qintai community, the survey sample in this study is mainly concentrated in the 26–40 age group, with relatively limited coverage of drivers aged 50 and above. Older drivers differ significantly from younger cohorts in terms of technology acceptance, travel purposes and frequency, as well as parking behaviors. For example, they tend to place greater emphasis on safety and reliability, make trips primarily for medical care and daily activities, and often occupy parking spaces for extended periods due to low vehicle usage. Neglecting this group may lead to an overestimation of residents’ acceptance of SAVs and an underestimation of persistent parking demand in aging communities, thereby affecting the representativeness of the findings and the generalizability of policy recommendations.

The parking demand measurement model mainly incorporates the core variables such as residents’ car ownership, time delay rate, and visitor traffic flow, but does not adequately cover the external intervention factors, such as the policy subsidies for the promotion of SAV, the adjustment of traffic restriction policy, and the speed of technological iteration, and also multidimensional potential influencing factors. The former may directly alter residents’ travel preferences, while the latter encompasses dimensions such as community public transportation accessibility like bus stop density, subway reachability, household income levels which influencing private vehicle purchases and willingness to pay for SAV usage, and family structure like presence of preschool-aged children or mobility-impaired members. This study prioritized validating the foundational mechanism of “core variable—demand change,” resulting in selective inclusion of variables. This approach may lead to an incomplete explanation of the driving logic behind SAV adoption and parking demand shifts, making it challenging to capture the interactions among multiple factors.

Both qualitative and quantitative analyses carry risks of data bias. At the qualitative level, residents’ acceptance and trust of SAVs are measured by a Likert scale are heavily influenced by subjective attitudes. at the quantitative level, parameters such as the rate of change of car ownership and the time lag rate are estimated based on the self-reported questionnaire data, which are prone to two types of bias: First, social desirability bias, respondents may overstate acceptance of SAVs to align with technological trends. Second, recall bias, discrepancies between recollections of past parking durations and travel frequencies and actual occurrences. In addition, the substitution coefficient of SAVs for traditional vehicles relies on existing literature and is not calibrated with the actual operation data of the case communities. Furthermore, the SAV substitution coefficient for conventional vehicles relies solely on existing literature values without calibration using actual operational data from case communities. More critically, key model parameters, such as time-of-day dwell rates and SAV substitution coefficients have not undergone sensitivity testing. This prevents quantifying the impact of parameter fluctuations within reasonable ranges on parking demand forecasts, thereby reducing the robustness and reliability of the model’s conclusions.

The design of the staged strategy focuses on the macro direction, but lacks sufficient exploration of specific details such as integration plans for SAV parking facilities with traditional parking lots and usage guidance measures for residents of different age groups; meanwhile, It also fails to incorporate potential risk analyses in SAV promotion, including travel safety risks caused by technical failures and the impact of SAV proliferation on community employment structures. Future research should integrate refined cost calculations and risk assessment models to enhance the strategy’s operational feasibility.

Furthermore, this study’s assessments of SAV acceptance and parking demand are based on the current stage of technological development and the policy environment, constituting static projections. Should subsequent developments accelerate—such as the commercialization of fully autonomous driving, policy shifts like prioritizing SAV road rights or introducing parking incentives, or changes in resident travel habits—actual adoption rates and parking demand may significantly deviate from current projections. Therefore, the study’s conclusions have temporal limitations. The current sample size of 135 cases is indeed somewhat insufficient for the research findings. Given the limited sample size of 135 cases at present, future periodic updates to the research sample are necessary to conduct data comparisons across short-term (1–3 years), medium-term (5–10 years), and long-term (10+ years) horizons, dynamically refining the demand forecasting model and strategy recommendations.

To address the above issues, future research can be enhanced through multidimensional improvements: First, expand the scope of case studies to include newly built gated communities, low-density suburban communities, and samples from different countries, thereby strengthening the universality of conclusions and their international reference value. Second, optimize survey design by increasing sample size and improving coverage of the elderly population, while integrating real-time traffic flow data from community parking lots to reduce subjective and memory biases inherent in self-reported questionnaires. Third, refine the modeling framework by incorporating variables such as public transportation accessibility and resident income, while adding sensitivity tests for key parameters to quantify the impact of parameter fluctuations on outcomes. Fourth, establish a dynamic tracking mechanism to conduct periodic near-, medium-, and long-term comparative studies. This approach should integrate technological advancements and policy adjustments to dynamically refine parking demand forecasts and strategy details, thereby enhancing the practical guidance value of the research.