The prioritized design elements for urban regeneration in the era of autonomous vehicles

ABSTRACT

This study derives design elements for urban regeneration in the autonomous vehicle era through literature review and interdisciplinary discussions and prioritizes the design elements via expert survey and hierarchical composite analysis. The derived design elements were organized into five regeneration categories ―environment and landscape (EL), transportation and technology (TT), resource and energy (RE), society and culture (SC), and economy and industry (EI)― each with six design elements and their priorities were analyzed in relation to ten urban spatial types classified under four criteria. The most prioritized element among the 13 composite weighted design elements was inclusive design (TT1), followed by smart infrastructure (TT3), pedestrian friendly street (EL4), public open space (EL3), and renewable energy (RE3) with regards to spatial types of parking only (H), transportation infrastructure (J), and lower floors (B). This study contributes to the growing body of research on renewal of vehicle-occupied space in the autonomous future, thus assisting in the practical implementation of sustainable urban regeneration.

Graphical Abstract

KEYWORDS:

• Autonomous vehicles
• urban regeneration
• design elements
• urban spatial types
• autonomous urbanism
• hierarchical composite analysis

1. Introduction

Cities have long undergone regeneration through adapting new mobility to past structures. Roads in medieval cities were built to accommodate the width of carriages, and Le Corbusier’s Ville Radieuse was planned based on automotive roads. The mass production of automobiles in the early 20th century caused traffic congestion in Manhattan, resulting in the construction of elevated railway for traffic relief. However, by the early 21st century, the elevated railway had been abandoned because of industrial shifts and was transformed into a public park, thus creating another new city landscape.

Today’s cities are facing the most complicated mobility environment. In 2020, Google’s self-driving app service, Waymo One, was launched publicly in Phoenix, Arizona, and WeRide has conveyed more than 140,000 passengers in Guangzhou. The Ministry of Land, Infrastructure and Transport of the Republic of Korea reported the successful demonstration of autonomous truck platooning on the country’s highways. Moreover, Honda announced their mass production of the world’s first level-three autonomous vehicle (AV) in 2021.

During this urban mobility transition, researchers have put forth optimistic and pessimistic scenarios for autonomous city from the following perspectives: digitalization (self-driving), sustainability (electric-charging), urbanization (shared-mobility) and market penetration. Previous researchers determined that one likely scenario of converting to fully automated and shared systems will increase opportunities for reshaping urban space within the next 30 years (Duarte and Ratti 2018; Fraedrich et al. 2019; Jang, Jang, and Song 2018; Jungwirth 2016; Milakis et al. 2017; Papa and Ferreira 2018; Stead and Vaddadi 2019; Townsend 2014; Zhang et al. 2015). Despite the great increase of research on various scenarios and policies for cities in the AV era (Gonzalez-Gonzalez, Nogués, and Stead 2020), there are only a few investigations on regeneration design strategies in accordance with urban space. Thus, this study derives design elements for sustainable urban regeneration and prioritizes the elements in relation to urban spatial types under the aforementioned scenario.

Given the necessity of urban regeneration in the autonomous future, in Section 2, ten objectives of urban regeneration derived through literature review were developed into five regeneration categories with six corresponding design elements via expert feedback, and the existing urban space was classified into ten spatial types. Section 3 describes the methods of expert survey, weighting analysis, and hierarchical composite weight analysis for prioritizing the derived design elements in relation to the spatial types. Section 4 presents the result of the analyses and suggests design approaches for each space type utilizing the prioritized design elements. Lastly, the four key findings of this study are summarized.

2. Derivation of design elements

The objectives of urban regeneration were derived via literature review and developed into five regeneration categories with corresponding six design elements through interdisciplinary discussions, and urban space was classified into ten spatial types.

2.1. Literature review

A wide range of studies on autonomous urbanism, including articles, reports, and design cases were reviewed for the following three aspects: 1) analysis and forecasting, 2) strategies and policies, or 3) planning and design, thus deriving a total of ten comprehensive objectives for urban regeneration in the AV era.

2.1.1. Analysis and forecasting

Much urban space is currently planned for vehicles, which are parked 95% of the time (Shoup 2005), could be freed up in the AV era (Cohen 2017; Zhang et al. 2015). Self-driving cars no longer must be parked in the same building where a driver is located. When a passenger exits, the AV can self-park outside the city center in an area where parking fees are relatively low or engage in shared mobility (Fraedrich et al. 2019). Thus, the demand for parking facilities in the city centers could decrease in the AV era (Chapin et al. 2016). The transition to shared mobility models will bring tangible reductions in current transportation infrastructure. In the case of Singapore, a scenario predicted a significant reduction of the current parking area from the current occupancy of 2.2% (15.8 square kilometers) of the country’s land area to 0.3% (2.2 square kilometers). As much as 86% of current parking spaces would be freed up or repurposed if a complete switch from private cars to shared mobility (Kondor et al. 2018). The areas for roads and parking facilities, which account for 65% and 45% of the downtown areas of Houston and Washington, DC, respectively, could be reduced by up to 90% of the existing area in the AV era (Glus, Rothman, and Iacobucci 2017). Moreover, 73% of the existing parking space in an office building located in the National Landing area will become unnecessary by 2040, freeing up approximately 55,000 m² of floor area (Barber, Carey, and Kang 2018).

If the existing road area decreased owing to fewer and narrower traffic lanes resulting from the improved accuracy of self-driving technology, the sidewalk widths of the current 6.63 million kilometers of U.S. roads could extend up to 9.1 meters (Jencek and Unterreiner 2018). Therefore, these areas could be reimagined as vibrant public open spaces supporting various outdoor activities beyond the frontages of adjacent buildings (Schlossberg et al. 2018). Combined with fully automated and electric charging facilities, typically smaller personal vehicles will reduce the current sizes of parking spaces and areas. Moreover, zero-emission and low noise level AVs are likely to be utilized as a part of interior space, affecting architectural planning and design (Jeong and Lee 2017; Nissan Europe 2016; Park and Yoon 2020; Toyota USA 2020).

However, the time freed up during autonomous travel between home and work could lead to a greater preference for living in the suburbs, leading to unprecedented urban sprawl (Papa and Ferreira 2018; Romem 2013). As the urban population gradually migrates to the suburbs, the density of existing downtown areas could decrease, reducing real estate values and creating difficulties in government tax management. Therefore, maintaining the existing density and size of these cities might become more difficult, possibly causing some cities to shrink or collapse (Larco 2018).

As a significant portion of the existing urban space occupied by vehicles (vehicle-occupied space, VOS) gradually reclaimed in the autonomous future, cities have to adapt and find solutions to integrate these new mobility services in efficient and sustainable ways (UITP 2020). How then should sustainable objectives be set for regenerating the built environment in the AV era?

2.1.2. Strategies and policies

In the AV era, it will be crucial to develop multifaceted strategies and policies that are suitable for cities and countries and that will ensure safe and inclusive built environment via shared, automated public transportation system (Dror et al. 2019; Gavanas 2019; Glus, Rothman, and Iacobucci 2017; González-González, Nogués, and Stead 2020). A future policy for the city of Seattle was established from six perspectives of justice, public safety, economic innovation, information management, and transportation technology (Corey et al. 2017), while the city of Toronto’s Smart City Comprehensive Plan was organized into six initiatives, including mobility, public sector, buildings and housing, sustainability, infrastructure, and digital innovation (SideWalk Labs 2019).

Leaders of the city of Somerville Massachusetts studied strategies for regenerating the existing parking spaces in the areas around Assembly Row and Union Square in relation to the future parking demand in those areas (Chin 2015), and Sasaki proposed equitable, safe, and transparent planning strategies for seven different regions of historic downtown, middle neighborhood, suburbs, airport, industrial, campus and greenfield in Greater Boston, imagining more livable urban space in the AV era (Mohamed et al. 2018). Additionally, leaders proposed planning strategies and frameworks for different street types to achieve safe, sustainable, equitable, and vibrant urban streetscapes in the autonomous future (Kisner et al. 2019; Oh et al. 2017; Riggs, Appleyard, and Johnson 2020).

2.1.3. Planning and design

It is important to examine design studies on built environment in the autonomous future that describes opportunities for adaptively reusing the reclaimed VOS, including roads and parking areas, in dense urban areas (Baumgardner 2015; Chapin et al. 2016; Cohen 2017; Fraedrich et al. 2019; Knowles, Ferbrache, and Nikitas 2020; Yigitcanlar, Wilson, and Kamruzzaman 2019). In New York, for instance, transforming parking areas could allow redevelopment of 2.5 million residential units, 290 million square meters of workplace, or a green space 10 times larger than Central Park (KPF 2019).

Researchers have proposed various streetscapes in the AV era, transforming reclaimed road spaces into vibrant pedestrian-oriented spaces that accommodate socio-cultural activities as well as environmentally friendly spaces that feature green infrastructure (Jencek and Unterreiner 2018; Tierney and Ruhl 2017). Others proposed integrating programmed spatial modules into reclaimed curbsides in New York City, resulting in citywide expansion and aggregation on environmental and cultural demand (FXFowle and Sam Schwartz 2017).

Other researchers examined concepts for adaptive reuse of existing office, residential, and civic facilities in the autonomous future (Jeong and Lee 2017; Park and Yoon 2020) as well as a programmatic design toolkit for a variable conversion of the existing office based on predictions of future parking demand (Barber, Carey, and Kang 2018). Additionally, some have presented concepts of smart cities wherein AVs become mobile power supplies responding to energy consumption in buildings or on the city grid, emphasizing the importance of intelligent amplification and future mobility for all (Nissan Europe 2016; Toyota USA 2020).

2.1.4. Literary derived objectives

Table 1 summarizes ten urban regeneration objectives derived through literature review on the AV era: ecological environment, public space, transport system, smart infrastructure, logistics and resources, renewable energy, social welfare, cultural lifestyle, local economy, and adaptive reuse.

Table 1. Objectives for urban regeneration in the AV era.

Type*	Authors	Ecological environment	Public space	Transportation system	Smart infrastructure	Resource logistics	Renewableenergy	Socialwelfare	Culture lifestyle	Local economy	Adaptivereuse
		18	23	27	27	12	14	15	23	22	16
AF	Milakis et al. (2017)	○		○	○		○			○
	Papa and Ferreira (2018)	○	○	○	○		○		○	○	○
	Stead and Vaddadi (2019)	○	○	○	○			○	○	○	○
	Yigitcanlar, Wilson, and Kamruzzaman (2019)	○	○	○	○	○	○	○	○	○	○
SP	Corey et al. (2017)		○	○	○	○		○
	Glus, Rothman, and Iacobucci (2017)			○	○			○	○	○	○
PD	Hyundai (2020)			○	○	○	○	○	○	○
	FXFowle and Sam Schwartz (2017)	○	○	○	○	○			○	○
	Nissan Europe (2016)	○	○	○	○		○
	Toyota USA (2020)	○	○	○	○	○	○		○	○	○
AF+SP	González-González, Nogués, and Stead (2020)	○	○	○	○	○	○	○	○	○	○
	Rothnie, Uffer, and Ghojeh (2016)		○	○	○				○	○
AF+PD	Barber, Carey, and Kang (2018)			○	○	○	○		○		○
	Baumgardner (2015)	○	○	○	○			○	○	○	○
	Fraedrich et al. (2019)	○	○	○	○		○		○	○	○
	Jencek and Unterreiner (2018)	○	○	○	○				○	○
	Jeong and Lee (2017)		○	○	○	○		○	○	○	○
	KPF (2018)		○	○	○		○		○	○	○
	Skinner and Bidwell (2016)	○	○	○	○	○	○	○	○	○
	Tierney and Ruhl (2017)	○	○	○	○
SP+PD	Audi Urban Future Initiative (2015)		○	○	○		○		○	○	○
	Mohamed et al. (2018)	○	○	○	○	○		○	○	○	○
	Riggs, Appleyard, and Johnson (2020)	○	○	○	○			○	○		○
AF+SP+PD	Kisner et al. (2019)	○	○	○	○	○	○	○	○	○
	Knorr (2018)	○	○	○	○			○	○	○	○
	Oh et al. (2017)		○	○	○			○	○	○	○
	SideWalk Labs (2019)	○	○	○	○	○	○	○	○	○

* Literature type: analysis and forecasting (AF), strategies and policy (SP), planning and design (PD)

2.2. Derivation of design elements

The ten objectives derived via literature review were organized into five regeneration categories and developed into six corresponding design elements for each category via expert discussions (Figure 1). To develop design elements in relation to urban space, smart technology, and social policy that were emphasized in the literature, three professionals in field of architecture and urban design, information and communication technology (ICT), and social science discussed the topic virtually in a group and in-person one-on-one. In accordance with experts’ feedback, the five regeneration categories each comprised six design elements to allow for subsequent weight analysis under the same conditions.

Figure 1. Derivation of design elements via literature review and expert feedback.

2.3. Derived design elements

2.3.1. Regeneration categories

The five regeneration categories derived through literature review and expert feedback previously are as follows.

1. Environment and landscape (EL): For a healthy coexistence between humans and nature, vehicle-occupied space (VOS) in urban area is utilized to improve air quality and water system and restore biodiversity for a pedestrian-friendly green city.

2. Transportation and technology (TT): To achieve inclusive urban environment in both physical and digital aspects, advanced ICT technology is applied to expand smart infrastructure and shared mobility service.

3. Resource and energy (RE): To achieve self-sufficient communities, food, water, and energy resources are provided through urban agricultural, renewable energy, and rainwater management systems transformed from current VOS.

4. Society and culture (SC): To improve social and cultural value in community, various social overhead capital including civic, medical, educational, and athletic facilities, as well as affordable housing, is expanded through mixed use of current VOS.

5. Economy and industry (EI): To improve the economic status of the communities, measures to recover vitality in old downtowns and expand industrial opportunities are implemented via utilizing current VOS.

2.3.2. Design elements

Table 2 describes design elements for the five regeneration categories developed and verified through expert brainstorming and discussion.

Table 2. Design elements for urban regeneration in the AV era.

Category	Design elements		Objectives
Environment and landscape (EL)	EL 1	Urban forest	Improve air quality and expand green areas via transforming urban vehicle-occupied space (VOS) to forest
	EL 2	Green infrastructure	Restore biodiversity and manage rainwater by converting VOS to biotopes and green infrastructure
	EL 3	Public open space	Increase the number of collective spaces for recreational use at different scales via planning VOS into parks and plazas
	EL 4	Pedestrian friendly street	Provide generous pedestrian zones and improve their conditions in relation to building frontage via reclaiming curbsides
	EL 5	Elevated connection	Enhance urban connectivity via repurposing the existing viaducts to elevated pathways
	EL 6	Resilient waterfront	Increase buffer capacity of waterfront and improve connections between city and the waterfront via reclaiming VOS
Transportation and technology (TT)	TT 1	Inclusive design	Ensure a safe and convenient urban environment both physically and digitally for all
	TT 2	Mobility lane	Plan protected mobility lanes for bicycles, personal transporters, and autonomous vehicles on the existing roads and ensure flexible operation of them on demand
	TT 3	Smart infrastructure	Improve the existing transportation infrastructure to feature intelligent automation on AVs’ parking, charging, sensoring, and securing process
	TT 4	Shared mobility service	Enhance public transportation network via establishing mobility-as-a-service in blind spots or last miles
	TT 5	Smart mobility hub	Upgrade the existing stations and terminals as smart mobility hubs and redevelop them as mixed-use facilities
	TT 6	AV as a movable space	Plan a new interior or exterior space on demand via attaching a 3d-printed AV to a building or clustering with other AVs
Resource and energy (RE)	RE 1	Green retrofitting	Manage building energy consumption and make an efficient use of VOS in a building via remodeling or relocating mechanical, electrical, and plumbing equipment
	RE 2	Green building envelope	Mitigate heat Island effect and manage rainwater runoff and supply on a building by converting a building envelope into green system
	RE 3	Renewable energy	Reduce a building’s energy consumption via expansion of renewable energy systems including photovoltaic and solar thermal on reclaimed VOS
	RE 4	Urban agriculture	Supply food resources for a community by expanding urban agriculture such as vertical production and rooftop farms on reclaimed VOS
	RE 5	AV as movable power	Establish on demand movable power supply system via sharing AVs’ electric power with a building or community
	RE 6	Demolition waste recycling	Recycle waste generated from the existing building and infrastructure demolition for smart transportation infrastructure construction
Society and culture (SC)	SC 1	Civic cultural use	Repurpose VOS in civic or office buildings into cultural and athletic uses
	SC 2	Independent living	Plan healthcare facilities and independent living communities for both senior and disabled on reclaimed VOS
	SC 3	School community center	Reuse VOS in educational facilities as mixed-use or cultural programs for community in response to the decline of school age population
	SC 4	Public housing	Provide more housing opportunities for vulnerable social groups by redeveloping the existing public parking lots and garages into public housing
	SC 5	Historic tourism	Improve tourism facilities and public realm in historic districts via utilizing VOS
	SC 6	Emergency management	Expand support facilities in preparation for both natural and social disaster relief via utilizing VOS
Economy and industry (EI)	EI 1	Retail corridor	Create vibrant neighborhood retail streets by utilizing curbsides and shared spaces
	EI 2	Old downtown revival	Revitalize local economy via converting of VOS into shared accommodations and improve the traditional market environment in old downtowns
	EI 3	Automated fulfillment	Expand automated fulfillment centers via reusing VOS or incorporating them into the existing transportation infrastructure such as interchange or underneath viaduct
	EI 4	Creative co-workspace	Plan diverse and flexible co-workspace and common areas via repurposing VOS of the existing offices
	EI 5	Living lab	Expand living labs for research and development (R&D) institutes at different scales and locations via utilizing VOS
	EI 6	Tech industrial complex	Plan and relocate facilities for new technology industrial companies via utilizing VOS of the existing industrial complex

2.4. Urban spatial types

As the demand for current vehicle-occupied space (VOS) gradually decreases, regeneration design strategies for these spaces are essential. Therefore, VOS was classified into ten spatial types under the following criteria: 1) division of drive and park areas, 2) presence of reusable structure(s), 3) connection between interior and exterior, and 4) property ownership (Figure 2) for subsequent analysis to prioritize the derived design elements in relation to the urban spatial types.

Figure 2. Urban spatial type classification.

The existing urban space was classified into driving and parking areas; although there are legal classifications for both, they were classified for this study according to the presence of reusable structures. Because it is difficult to limit all case of spatial structures in terms of their scales and usage, the areas with structures were classified according to their connection between interior and exterior, and areas without structures were classified considering publicness according to property ownership. Because the parking only areas in downtown areas are highly likely to be repurposed, they were classified as an independent spatial type regardless of the presence of structures. Additionally, automotive facilities other than parking were classified separately because they can be redeveloped with taking the advantage of their locational conditions, such as visibility and accessibility, when their original demand decreases. In Figure 2, basements (A) are existing below-grade parking areas, including vehicle ramps, parking spaces, and aisles in a building. Lower floors (B) include existing at-grade drop-off areas, vehicle entrances and ramps, loading areas, and lower floors in a building. Middle floors (C) in a building might house AVs in the future, allowing their direct access. Likewise, upper floors (D) in a building, including rooftop parking areas, could have a similar impact on AVs. Non-building areas within a property boundary, including on-site surface parking or vacant areas, were classified as open space in private property (E), while open space in public property (F) include sidewalks and non-building areas within a public realm. Streets (G) is limited to urban streets, including curbside, on-street parking, shared spaces, alleys, and back streets. A surface or structured parking area designated only for parking was classified as parking only (H), whereas automotive facilities (I) include gas stations, car washes, repair shops, or automotive facilities other than parking. Lastly, highways, roundabouts, viaducts, tunnels, interchanges, and other than streets (G) were grouped as transportation infrastructure (J).

3. Survey and weight analysis

A questionnaire and weight analysis were applied to identify the importance of each regeneration category, design element, and spatial type and to prioritize design elements in relation to the spatial types. The expert survey and hierarchical composite weight analysis method for priority screening of the derived design elements and spatial types in Section 2 are described below.

3.1. Expert survey

Professionals in architecture, landscape architecture, and urban design were surveyed regarding the weight they assigned to each design element and spatial type. The respondents were from industry, government, and academia: employers or employees in planning and design, engineering, and construction private enterprises; government officials; and professors and researchers. The survey, created by Google Forms, was distributed to the survey participants via e-mail or instant messenger. The contents of the survey were explained through phone calls or in-person interviews for those who sought further explanation. The survey was implemented from December 14 to 23, 2020, and from 157 surveys distributed, 103 responses were received, with a 65.6% response rate.

According to the responses (Figure 3), the largest proportion (63.2%) of all respondents were in industry, followed by academia (28.4%) and government (8.4%), 71.6% were in planning and design, 55.8% were master’s degree holders, 34.7% had 10 to 20 years of experience, and 79.8% were residents from Seoul metropolitan area. The initial target of response rate was not achieved, however the survey result reflects global opinions some extent, including New York, Paris, Berlin, and Sydney.

Figure 3. Survey respondents’ attributes.

The survey was hierarchically organized into four sections, namely, 1) backgrounds of AVs and urban mobility, 2) weight of regeneration category and design elements, 3) weight of urban spatial types, and 4) respondents’ attributes, with a total of 74 multiple-choice and Likert scale selection questions. The weight of each item was scored on a 5-point Likert scale ranging from not at all important (0) to very important (4). To measure the reliability (internal consistency) of the 103 collected questionnaires, scale analysis was performed using IBM SPSS. A total of 95 effective samples were extracted, and Cronbach’s alpha was calculated for each section, yielding values ranging between 0.7 and 0.8 (Table 3), which exceeded the criterion of 0.6¹ (George and Mallery 2003).

Table 3. Reliability test results.

Survey sections		Cronbach’s alpha	Number of items
Regeneration category		0.720	5
Design elements	Environment and landscape (EL)	0.716	6
	Transportation and technology (TT)	0.736	6
	Resource and energy (RE)	0.734	6
	Society and culture (SC)	0.739	6
	Economy and industry (EI)	0.794	6
Urban spatial types		0.800	10

3.2. Weight analysis

A hierarchical composite analysis model (Figure 4) was designed for the weight analysis of the survey results. Each weight of regeneration category ($W_x$), design element ($W_{yl}$), and spatial weight ($W_s$) was derived with Formulas 1(1) $A=n^{-1}\cdot \sum _{k=0}^{4}k\cdot n_k\text{break}k=\hspace{0.17em}0,\hspace{0.17em}1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}4\hspace{0.17em}\left(Likertscalescore\right)\hspace{0.17em}$(1) and 2(2) $W=\frac{A}{SA}$(2) in order. The total weight ($T{W}_{zl}$), calculated from the product of $W_x$ and corresponding $W_{yl}$, was added to $W_s$ to yield the composite weighted ($CW$) design elements for urban regeneration in the AV era. In the analysis, the weight given to each section’s weight was compared with the respondents’ sector.

Figure 4. Hierarchical composite analysis model.

A 5-point Likert score ($A$) for each group was calculated as the sum of all products ($\sum _{k=0}^{4}k\cdot n_k$) of the Likert score ($k$) and the number of samples per corresponding scale ($n_k$) divided by the total number of samples ($n$; Formula 1). The weight ($W$) of each item was calculated by dividing the scale average ($A$) by the sum of the scale averages ($SA$) of all groups (Formula 2(2) $W=\frac{A}{SA}$(2) ):

(1) $A=n^{-1}\cdot \sum _{k=0}^{4}k\cdot n_k\text{break}k=\hspace{0.17em}0,\hspace{0.17em}1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}4\hspace{0.17em}\left(Likertscalescore\right)\hspace{0.17em}$(1)

(2) $W=\frac{A}{SA}$(2)

Each scale average of the five regeneration categories ($A_x$) was calculated using Formula 1(1) $A=n^{-1}\cdot \sum _{k=0}^{4}k\cdot n_k\text{break}k=\hspace{0.17em}0,\hspace{0.17em}1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}4\hspace{0.17em}\left(Likertscalescore\right)\hspace{0.17em}$(1) , and each category weight ($W_x$) was derived by dividing the corresponding category’s scale average ($A_x$) by the sum of all scale category averages (Formula 3(3) $W_x=\frac{A_x}{A_{EL}+A_{TT}+A_{RE}+A_{SC}+A_{EI}}\text{break}x\mathit{=}EL\mathit{,}TT\mathit{,}RE\mathit{,}SC\mathit{,}EI$(3) ):

(3) $W_x=\frac{A_x}{A_{EL}+A_{TT}+A_{RE}+A_{SC}+A_{EI}}\text{break}x\mathit{=}EL\mathit{,}TT\mathit{,}RE\mathit{,}SC\mathit{,}EI$(3)

Each scale average of the six elements ($A_{yl}$) in the corresponding category was calculated by Formula 1(1) $A=n^{-1}\cdot \sum _{k=0}^{4}k\cdot n_k\text{break}k=\hspace{0.17em}0,\hspace{0.17em}1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}4\hspace{0.17em}\left(Likertscalescore\right)\hspace{0.17em}$(1) , and each element weight ($W_{yl}$) was calculated by dividing the scale average ($A_{yl}$) by the sum of all scale averages of the elements, as in Formula 4(4) $W_{yl}=A_{yl}\cdot {\left(\sum _{l=1}^6{A}_{yl}\right)}^{-1}\text{break}l=1,2,3,4,5,6\text{break}y=EL,TT,RE,SC,EI$(4) . For instance, the scale average of EL1 ($A_{EL1}$) was calculated by the sum of all products ($\sum _{k=0}^{4}k\cdot n_k$) of Likert score ($k$) and the corresponding number of samples ($n_k$) was divided by the total number of samples ($n$). The weight of EL1 ($W_{EL1}$) was derived by dividing the scale average of EL1 ($A_{EL1}$) by the sum of all scale averages from EL1 to EL6.

(4) $W_{yl}=A_{yl}\cdot {\left(\sum _{l=1}^6{A}_{yl}\right)}^{-1}\text{break}l=1,2,3,4,5,6\text{break}y=EL,TT,RE,SC,EI$(4)

Each design element’s total weight ($T{W}_{zl}$) was derived by multiplying its corresponding category weight ($W_x$) and element weight ($W_{yl}$), as shown in Formula 5$\begin{array}{rl}& T{W}_{zl}=W_x\cdot W_{yl}\\ & x=y=z=EL,TT,RE,SC,EI(5)\end{array}$. For instance, EL1’s total weight ($T{W}_{EL1}$) was the product of the EL category weight ($W_{EL}$) and the EL1 element weight ($W_{EL1}$).

$\begin{array}{rl}& T{W}_{zl}=W_x\cdot W_{yl}\\ & x=y=z=EL,TT,RE,SC,EI(5)\end{array}$

The scale averages for the 10 spatial types ($A_s$) were calculated using Formula 1(1) $A=n^{-1}\cdot \sum _{k=0}^{4}k\cdot n_k\text{break}k=\hspace{0.17em}0,\hspace{0.17em}1,\hspace{0.17em}2,\hspace{0.17em}3,\hspace{0.17em}4\hspace{0.17em}\left(Likertscalescore\right)\hspace{0.17em}$(1) , and each spatial weight ($W_s$) was calculated by dividing $A_s$ by the sum of scale averages of all 10 spatial types (Formula 6(6) $W_s=A_s\cdot {\left(\sum _{m=a}^j{A}_m\right)}^{-1}$(6) ):

(6) $W_s=A_s\cdot {\left(\sum _{m=a}^j{A}_m\right)}^{-1}$(6)

In the final step of the hierarchical composite analysis model, the composite weights ($CW$) were obtained from the design element’s total weight ($T{W}_{zl}$) and spatial weight ($W_s)$ as follows. First, the number of design elements and space types applicable to one another were counted ($c$), and the total weights ($T{W}_{zl}$) and spatial weights ($W_s$) were converted into points ($p$) in descending order. Each point ($p_{zl}$) of total weight ($W_{zl}$) from 30 to 1 was added to the number ($c_{zl}$) of applicable design elements from 10 to 1, resulting in a perfect score of 40 ($p_{zl}+c_{zl}$). Likewise, each point ($p_s$) of spatial weight ($W_s$) from 10 to 1 was added to the number ($c_s$) of applicable space types from 30 to 1, resulting in a perfect score of 40 ($p_s+c_s$). The composite weight ($CW$) was calculated using Formula 7(7) $CW=(p_{zl}+c_{zl})+(p_s+c_s)$(7) . Based on a perfect score of 80, the composite weight ($CW$) was prioritized into three levels: 1) top priority ($C{W}_{tp}$) score greater than 64 (20%), 2) middle priority ($C{W}_{mp}$) between 64 (20%), and 3) 48 (40%), and low priority ($C{W}_{lp}$) below 48 (40%; Formula 8(8) $C{W}_{tp}\ge 64>C{W}_{mp}\ge 48>C{W}_{lp}$(8) ).

(7) $CW=(p_{zl}+c_{zl})+(p_s+c_s)$(7)

(8) $C{W}_{tp}\ge 64>C{W}_{mp}\ge 48>C{W}_{lp}$(8)

4. Result of weight analyses

In Section 4, weight analyses results, prioritization of the design elements in accordance to spatial type, and design approach through the application of the prioritized design elements are described.

4.1. Category weight

In the regeneration category weight analysis, all respondents rated transportation and technology (TT) as the most important category, followed by environment and landscape (EL) and economy and industry (EI). In contrast, the industry and academia respondents considered society and culture (SC) the least important category, the government workers ranked resource and energy (RE) (Table 4).

Table 4. Category weight analysis results.

Regeneration category	Overall		Industry		Government		Academia
	$W_x$ *	Rank	$W_x$	Rank	$W_x$	Rank	$W_x$	Rank
Environment and landscape (EL)	0.202	2	0.202	2	0.207	2	0.200	2
Transportation and technology (TT)	0.219	1	0.216	1	0.214	1	0.225	1
Resource and energy (RE)	0.196	4	0.197	4	0.179	5	0.198	3
Society and culture (SC)	0.187	5	0.185	5	0.200	3	0.186	5
Economy and industry (EI)	0.197	3	0.199	3	0.200	3	0.191	4

* $W_x$= category weighting factor

4.2. Element weight

The element weight analysis of the respondents showed that pedestrian friendly street (EL4), inclusive design (TT1), renewable energy (RE3), public housing (SC4), and living lab (EI5) were the most important elements, and resilient waterfront (EL6), AV as moveable space (TT6), demolition waste recycling (RE6), historic tourism (SC5), and automated fulfillment (EI3) were the least important in their respective categories (Table 5).

Table 5. Element weight analysis results.

Design elements		Overall		Industry		Government		Academia
		$W_{yl}$*	$Rank$	$W_{yl}$	$Rank$	$W_{yl}$	$Rank$	$W_{yl}$	$Rank$
EL 1	Urban forest	0.176	3	0.178	3	0.159	4	0.178	3
EL 2	Green infrastructure	0.159	4	0.157	4	0.165	3	0.160	4
EL 3	Public open space	0.185	2	0.182	1	0.195	1	0.186	2
EL 4	Pedestrian friendly street	0.185	1	0.182	1	0.177	2	0.194	1
EL 5	Elevated connection	0.150	5	0.148	6	0.159	4	0.152	5
EL 6	Resilient waterfront	0.145	6	0.153	5	0.146	6	0.128	6
TT 1	Inclusive design	0.185	1	0.183	1	0.183	1	0.188	1
TT 2	Mobility lane	0.175	3	0.173	2	0.175	3	0.178	3
TT 3	Smart infrastructure	0.176	2	0.172	3	0.182	2	0.180	2
TT 4	Shared mobility service	0.164	4	0.165	4	0.149	5	0.167	4
TT 5	Smart mobility hub	0.158	5	0.165	4	0.156	4	0.144	5
TT 6	AV as a movable space	0.144	6	0.143	6	0.149	5	0.144	5
RE 1	Green retrofitting	0.175	2	0.176	1	0.171	2	0.176	3
RE 2	Green building envelope	0.171	3	0.174	2	0.158	4	0.170	4
RE 3	Renewable energy	0.176	1	0.173	3	0.190	1	0.180	1
RE 4	Urban agriculture	0.157	5	0.163	4	0.158	4	0.143	6
RE 5	AV as movable power	0.168	4	0.162	5	0.171	2	0.178	2
RE 6	Demolition waste recycling	0.153	6	0.153	6	0.152	6	0.154	5
SC 1	Civic cultural use	0.167	3	0.167	4	0.151	5	0.173	3
SC 2	Independent living	0.175	1	0.172	1	0.189	1	0.177	2
SC 3	School community center	0.164	5	0.168	2	0.170	3	0.155	5
SC 4	Public housing	0.172	2	0.168	2	0.176	2	0.179	1
SC 5	Historic tourism	0.156	6	0.162	6	0.151	5	0.145	6
SC 6	Emergency management	0.167	4	0.163	5	0.164	4	0.171	4
EI 1	Retail corridor	0.168	2	0.169	3	0.149	6	0.171	2
EI 2	Old downtown revival	0.166	4	0.167	4	0.170	2	0.162	5
EI 3	Automated fulfillment	0.163	6	0.158	6	0.163	5	0.173	1
EI 4	Creative co-workspace	0.166	4	0.163	5	0.170	2	0.171	2
EI 5	Living lab	0.170	1	0.173	1	0.170	2	0.165	4
EI 6	Tech industrial complex	0.167	3	0.170	2	0.177	1	0.158	6

* $W_{yl}$= element weighting factor

4.3. Total weight

The total weight ($T{W}_{zl}$) analysis for respondents overall showed that inclusive design (TT1) was the most crucial of the 30 design elements, followed by smart infrastructure (TT3), mobility lane (TT2), pedestrian friendly street (EL4), and public open space (EL3). Among the top 10 (33%) total weighted design elements, 5 were included in TT, 3 in EL, and 2 in RE (Table 6).

Table 6. Total weight analysis results.

Design elements		Overall		Industry		Government		Academia
		$T{W}_{zl}$*	$Rank$	$T{W}_{zl}$	$Rank$	$T{W}_{zl}$	$Rank$	$T{W}_{zl}$	$Rank$
EL 1	Urban forest	0.036	7	0.036	6	0.033	16	0.036	7
EL 2	Green infrastructure	0.032	21	0.032	20	0.034	9	0.032	20
EL 3	Public open space	0.037	5	0.037	4	0.040	1	0.037	6
EL 4	Pedestrian friendly street	0.037	4	0.037	4	0.037	6	0.039	4
EL 5	Elevated connection	0.030	27	0.030	29	0.033	16	0.031	24
EL 6	Resilient waterfront	0.029	29	0.031	25	0.030	24	0.026	30
TT 1	Inclusive design	0.040	1	0.034	1	0.040	2	0.042	1
TT 2	Mobility lane	0.038	3	0.037	2	0.038	5	0.040	3
TT 3	Smart infrastructure	0.038	2	0.037	3	0.039	3	0.041	2
TT 4	Shared mobility service	0.036	6	0.036	7	0.032	20	0.038	5
TT 5	Smart mobility hub	0.034	9	0.036	7	0.033	15	0.032	17
TT 6	AV as a movable space	0.031	22	0.031	26	0.032	20	0.032	17
RE 1	Green retrofitting	0.034	10	0.035	9	0.031	22	0.035	10
RE 2	Green building envelope	0.033	11	0.034	11	0.028	28	0.034	11
RE 3	Renewable energy	0.035	8	0.034	12	0.034	14	0.036	8
RE 4	Urban agriculture	0.031	26	0.032	17	0.028	28	0.028	28
RE 5	AV as movable power	0.033	15	0.032	18	0.031	22	0.035	9
RE 6	Demolition waste recycling	0.030	28	0.030	28	0.027	30	0.030	25
SC 1	Civic cultural use	0.031	23	0.031	24	0.030	25	0.032	19
SC 2	Independent living	0.033	16	0.032	19	0.038	4	0.033	14
SC 3	School community center	0.031	25	0.031	22	0.034	13	0.029	27
SC 4	Public housing	0.032	19	0.031	22	0.035	8	0.033	12
SC 5	Historic tourism	0.029	30	0.030	30	0.030	25	0.027	29
SC 6	Emergency management	0.031	24	0.030	27	0.033	18	0.032	21
EI 1	Retail corridor	0.033	13	0.034	14	0.030	27	0.033	15
EI 2	Old downtown revival	0.033	17	0.033	15	0.034	10	0.031	23
EI 3	Automated fulfillment	0.032	20	0.031	21	0.033	19	0.033	13
EI 4	Creative co-workspace	0.033	17	0.033	16	0.034	10	0.033	15
EI 5	Living lab	0.033	12	0.034	10	0.034	10	0.031	22
EI 6	Tech industrial complex	0.033	14	0.034	13	0.036	7	0.030	26

* $T{W}_{zl}$= total weighting factor

4.4. Spatial weight

The spatial weight ($W_s$) analysis showed that lower floors (B) were the most important space type, followed by parking only (H), transportation infrastructure (J), and basements (A). In contrast, middle floors (C) were the least important space type alongside upper floors (D) in the autonomous future. Particularly, respondents in academia regarded open space in public (F) as the most crucial space type, unlike in the other two sectors (Table 7).

Table 7. Spatial weight analysis results.

Spatial types		Overall		Industry		Government		Academia
		$W_s$*	$Rank$	$W_s$	$Rank$	$W_s$	$Rank$	$W_s$	$Rank$
A	Basements	0.110	4	0.110	3	0.105	7	0.112	3
B	Lower floors	0.116	1	0.116	1	0.120	1	0.115	2
C	Middle floors	0.052	10	0.052	10	0.060	10	0.050	10
D	Upper floors	0.072	9	0.073	9	0.064	9	0.074	9
E	Open space in private	0.102	8	0.107	7	0.094	8	0.095	8
F	Open space in public	0.108	7	0.103	8	0.109	5	0.118	1
G	Streets	0.110	4	0.110	5	0.112	3	0.110	4
H	Parking only	0.112	2	0.112	2	0.116	2	0.109	6
I	Automotive facilities	0.108	6	0.108	6	0.109	5	0.108	7
J	Transportation infrastructure	0.111	3	0.110	3	0.112	3	0.110	4

* $W_s$= spatial weighting factor

4.5. Composite weight

Table 8 presents the prioritized design elements in relation to the ten spatial types for each level. Inclusive design (TT1), AV as a movable space (TT6), renewable energy (RE3), urban agriculture (RE4), AV as movable power (RE5), and living lab (EI5) were identified as design elements that are universally applicable to all spatial types. Additionally, all the design elements except mobility lane (TT2) could be applied to parking only (H) and automotive facility (I), whereas middle floors (C) and upper floors (D) had 14 and 16 applicable design elements, respectively, less than the other spatial types. Among the composite weighted design elements, top priority ($C{W}_{tp}$) accounted for 13 out of 30 (43.3%), 5 in TT, 3 each in EL and RE, 2 in EI, and none in SC. All 13 design elements were applicable to parking only (H), 10 to both transportation infrastructure (J) and lower floors (B), and 8 elements to automotive facilities (I).

Table 8. Composite weight analysis results.

● = $C{W}_{tp}\ge 64$ ◎ = $64>C{W}_{mp}\ge 48$ ○ = $48>C{W}_{lp}$				Spatial types
				Type	A	B	C	D	E	F	G	H	I	J
				$p_s$	7	10	1	2	3	4	7	9	5	8
Design elements			$p_{zl}$*	$c_s$$c_{zl}\ast \ast$	24	25	14	16	26	20	20	29	29	27
Environment and landscape (EL)	EL 1	Urban forest	24	6					◎	◎	◎	●	●	●
	EL 2	Green infrastructure	10	6					○	○	○	◎	◎	◎
	EL 3	Public open space	26	8	●	●			◎	◎	◎	●	●	●
	EL 4	Pedestrian friendly street	27	8	●	●			◎	◎	◎	●	●	●
	EL 5	Elevated connection	4	8	○	○			○	○	○	◎	○	○
	EL 6	Resilient waterfront	2	6					○	○	○	○	○	○
Transportation and technology (TT)	TT 1	Inclusive design	30	10	●	●	◎	◎	●	●	●	●	●	●
	TT 2	Mobility lane	28	3						◎	◎			●
	TT 3	Smart infrastructure	29	8	●	●			●	◎	●	●	●	●
	TT 4	Shared mobility service	25	8	◎	●			◎	◎	◎	●	●	●
	TT 5	Smart mobility hub	22	9	◎	●		◎	◎	◎	◎	●	●	●
	TT 6	AV as a movable space	9	10	◎	◎	○	○	◎	○	○	◎	◎	◎
Resource and energy (RE)	RE 1	Green retrofitting	21	6	◎	◎	○	○				●	◎
	RE 2	Green building envelope	19	6	◎	◎	○	○				◎	◎
	RE 3	Renewable energy	23	10	●	●	◎	◎	◎	◎	◎	●	●	●
	RE 4	Urban agriculture	5	10	○	◎	○	○	○	○	○	◎	◎	◎
	RE 5	AV as movable power	16	10	◎	◎	○	○	◎	◎	◎	●	◎	◎
	RE 6	Demolition waste recycling	3	5						○	○	○	○	○
Society and culture (SC)	SC 1	Civic cultural use	8	8	○	◎	○	○	○			◎	◎	◎
	SC 2	Independent living	13	8	◎	◎	○	○	◎			◎	◎	◎
	SC 3	School community center	5	8	○	○	○	○	○			◎	○	◎
	SC 4	Public housing	12	6		◎	○	○	○			◎	◎
	SC 5	Historic tourism	1	8	○	○			○	○	○	○	○	○
	SC 6	Emergency management	7	7	○	◎		○	○			◎	◎	◎
Economy and industry (EI)	EI 1	Retail corridor	18	8	◎	◎			◎	◎	◎	●	◎	◎
	EI 2	Old downtown revival	13	8	◎	◎			◎	◎	◎	◎	◎	◎
	EI 3	Automated fulfillment	11	6	○	◎			○			◎	◎	◎
	EI 4	Creative co-workspace	13	8	◎	◎	○	○	◎			◎	◎	◎
	EI 5	Living lab	19	10	◎	●	○	○	◎	◎	◎	●	◎	●
	EI 6	Tech industrial complex	17	8	◎	◎	○	○	◎			◎	◎	◎

* $p_{zl}$= total weight ($T{W}_{zl}$) converted into points, $p_s$= spatial weight ($W_s$) converted into points

** $C_{zl}$= number of applicable design elements, $C_s$= number of applicable spatial types

4.6. Design approach

The ideas of proposed for applying the prioritized design elements to each spatial type for regeneration in the AV era are as follows.

A building’s basements (A) will likely to be converted to public, educational, cultural, athletic, or mixed-use programs relocating or remodeling the existing basement mechanical, electrical, and plumbing equipment in the AV era. The reclaimed space could be used for a data center, for example, reducing building energy consumption by recycling waste heat from the data center equipment. In addition, vertical farming could be planned in interior spaces with irrigation by a rainwater harvesting system on the roof. Upper floors (D), especially rooftop parking areas, are suitable for accommodating renewable energy systems or urban farms, resulting in a potential energy and resource supply for a local community. Additionally, replacing the reclaimed rooftop parking areas with green space will alleviate the urban heat Island effect and reduce rainwater runoff. Converting the rooftop parking areas of terminals and stations in cities to aviation mobility stations could result in reinforced urban mobility hubs in the autonomous future. A zero-emission AV’s inclusive access to middle floors (C) in a building will result in spatial and programmatic linkage between the vehicle and interior space. Moreover, lower floors (B) will become collective and transitional spaces for people and AVs.

Open spaces in private (E) and public (F) property are likely to be redeveloped to accommodate various programs, including retail, community, culture, or green space interaction with adjacent buildings’ programs. Moreover, they could incrementally expand to the curbsides of streets (G) reclaimed in the autonomous future. With traffic safety facilities, including curbs and bollards to separate pedestrians and vehicles no longer needed, the existing roads could be replaced with vibrant and pedestrian friendly spaces, such as linear parks or urban forests interacting with building frontages.

Parking only (H) and automotive facilities (I) in urban center aeras are potential redevelopment sites for a range of mixed-use programs or public open spaces interacting with adjacent buildings’ programs. Transportation infrastructure (J) will enable wireless exchange of power and information among personal mobiles, AVs, and buildings via embedded sensor network systems. Moreover, the reclaimed space of highways and tunnels is a potential area for restored ecological habitats or planned renewable energy parks. Obsolete viaducts and bridges can be transformed into renewable energy infrastructure supplying local electric demand with mixed-use programs below.

5. Conclusion

In response to the increase of scenario and policy studies on cities in the AV era (González-González, Nogués, and Stead 2020; Papa and Ferreira 2018; Stead and Vaddadi 2019), this study was conducted to derive design elements for urban regeneration in the AV era and prioritize the design elements in accordance with urban spatial types. The findings and implications of this research are as follows.

First, a total of 30 design elements for urban regeneration were derived from multifaceted perspectives. Although previously there have been speculations on urban space in the AV era via various lenses, it was necessary to comprehend the different scopes and purposes of literature and examine it from the VOS regeneration perspective. Therefore, five regeneration categories, environment and landscape (EL), transportation and technology (TT), resource and energy (RE), society and culture (SC), and economy and industry (EI), and six corresponding design elements for each category were developed through a literature review and interdisciplinary discussions.

Second, the spatial scope for urban regeneration in the AV era was defined and analyzed in relation to the derived design elements. The existing urban space was classified into a total of ten spatial types based on four criteria to examine the applicability and priority of the derived design elements in relation to urban spatial types. However, additional studies are needed to determine detailed criteria for urban space classification because of the requirement for a broad consideration of site-specific conditions, such as usage and scale, required codes, and surrounding context.

Third, the weights of derived design elements and spatial types were analyzed quantitatively with expert surveys and weight analysis, thereby yielding more objective priorities for design elements. Respondents rated TT the most important category, SC the least. The most crucial design elements for each category were pedestrian friendly street (EL4), inclusive design (TT1), renewable energy (RE3), independent living (SC2), and living lab (EI5). Moreover, respondents rated lower floors (B) the most important space type, followed by parking only (H) and transportation infrastructure (J). However, to reach a broader social consensus, the scope of the survey should be expanded to a broader range of fields and the priorities of design elements updated accordingly.

Finally, prioritized design elements were derived through a hierarchical composite analysis model and divided into three levels. The upper two-thirds (8 out of 13) of the design elements (13 out of 30) were concentrated in the TT and EL categories. Inclusive design (TT1) received the highest priority, followed by smart infrastructure (TT3), public open space (EL3), pedestrian friendly street (EL4), and renewable energy (RE3), suggesting that urban regeneration in the AV era could aim to create safe and convenient urban environments for all of society on current urban VOS through a convergence of green and smart technologies. However, this regeneration will require public and stakeholder engagement and agreed-upon policies, such as mitigating of existing building codes and requirements, to increase the adaptive reuse opportunities of current urban VOS. For instance, reducing minimum parking or zoning changes can promote mixed-use redevelopment of parking only (H) and automotive facilities (I).

Because this study was conducted on the premise of several assumptions, the results cannot apply to all scenarios in the autonomous future. Thus, other assumptions and approaches could have resulted in more reliable research results, such as analyzing the priority of design elements for expanded scenarios, classifying urban spatial types in a specific city or region, or minimizing the ratio gap among survey respondent attributes for the higher reliability of the weight analysis results.

Nevertheless, this study has two main significances. First, the attempt of hierarchical composite analysis itself was important some extent, because the result of weighting analysis reflects multiple criteria which could be developed into an analysis model for other relevant studies. Second, the prioritized design elements in relation to urban spatial types could contribute to the studies on adaptive reuse or site renewal design guideline, thus contributing to the practical implementation of sustainable urban regeneration for the autonomous future.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Soongsil University Research Fund (New Professor Support Research) of 2020.

Notes on contributors

Chiyoung Park

Chiyoung Park is an assistant professor at Soongsil University School of Architecture and director of City Place Laboratory (www.c-plab.com). He received a Master of Landscape Architecture concentrated in Urban Design from the University of Pennsylvania and a Bachelor of Architecture from Soongsil University. He has professional experience from Gensler, Field Operations, PORT, PARKKIM, and Junglim where he practiced in a variety of mixed-use development and public realm projects from conceptual frameworks throughout all design phase delivery. He is the 3rd Prize winner of 2018 Seoul Jamsil Sports Complex Olympic Stadium Remodeling International Competition and the winner of 2016 Robert M. Hanna Prize in Design. He is currently a principal investigator of the research on Urban Regeneration Strategies and Design Guideline in the Autonomous Vehicle Era supported from the National Research Foundation of Korea.

Notes

¹ Cronbach’s alpha was interpreted under the following criteria: Excellent ($\alpha \ge 0.9)$, Good ($0.9\ge \alpha \ge 0.8),$ Acceptable ($0.8\ge \alpha \ge 0.7),$ Questionable ($0.7\ge \alpha \ge 0.6),$ Poor ($0.6\ge \alpha \ge 0.5),$ and Unacceptable ($0.5\ge \alpha )$.