Automating the first and last mile? Reframing the ‘challenges’ of everyday mobilities

Abstract In this article, we interrogate the utility of conceptualising the ‘first and last mile’ (FLM) as a ‘challenge’ to be addressed through automated and integrated mobility services. We critically engage with the concept through a design anthropological approach which takes two steps so as: to complicate literatures that construct the FLM as a place where automated, service-based and micro-mobility innovations will engender sustainable modal choices above individual automobility; and to demonstrate how people’s situated mobility competencies and values, shape social and material realities and future imaginaries of everyday mobilities. To do so, we draw on ethnographic research into everyday mobility practices, meanings and imaginaries in a suburban neighbourhood in Sweden. We show how locally situated mobilities both challenge the spatial and temporal underpinnings of the first and last mile concept, and resist universalist technology-driven automation narratives. We argue that instead of attempting to bridge gaps in seemingly linear journeys through automated systems, there is a need to account for the practices, tensions and desires embedded in everyday mobilities.


Introduction
In recent years the first and last mile (FLM) 'challenge', has become increasingly central to industry, consultancy and policy agendas for automated mobilities.In these dominant narratives, automated FLM solutions are seen to respond to a 'societal problem' whereby individuals use private cars to traverse the last miles in their everyday commutes due to lack of convenient public transport.The quest for FLM solutions echoes typical technological solutionist (Morozov 2013) approaches to creating and applying automated solutions to perceived social or societal problems, which endure in the mobilities sector (Quilty et al. 2022).
In this article, we interrogate the implications of formulating the FLM of passenger travel to and from home as a challenge which automated systems, technologies and services might solve.We confront dominant, universalist automation narratives with the mobility practices, tensions and desires that unfold in the actual, immediate time-spaces around the home.That is, we focus on how people navigate the spaces presented in FLM propositions.In doing so, we contribute to, and seek to complicate, recent literatures that focus on the first and last mile as a place for implementing automated, service-based and micro-mobility transport options in order to nudge people into more sustainable modes of travel other than using individually owned cars (Doody et al. 2022;Behrendt 2020;Cohen 2006).
FLM concepts are increasing part of the mobility solutions vocabulary.The 'last mile' concept has been used to refer to the final segment of freight transport journeys (Oliveira et al. 2017) and recently gained traction in the field of passenger mobility, where it is used to refer to gaps in transit coverage at the very beginning and end of commuter travel.First mile solutions are often based on an assumption that people use private cars to cover this part of their journey or indeed their entire journey, because they have insufficient public transport options, or lack incentives to use public transport when it is available (Lesh 2013;Mohiuddin 2021), especially in suburban contexts.Thus, within urban planning and transportation sustainability research it is suggested that improving access to transit networks or shared mobility options could reduce individual private car use and enable a shift towards shared mobility.
Possible FLM solutions proposed to replace privately owned cars for (sub)urban passenger journeys include: automated micromobility systems using lightweight, swift-paced micro-vehicles such as e-scooters, monowheels, bicycles or skateboards (Bahrami and Rigal 2022); as well as (Shared) Autonomous Vehicles (AVs).In this context, Mobility as a service (MaaS) (Wong, Hensher, and Mulley 2020) and Mobility on demand (MoD) systems are thought to offer increased efficiency and flexibility for AV sharing (Chong et al. 2013;Jones et al. 2023).
However, existing visions of automated mobility solutions are limited in several ways.First, they are often naïve in their technological-determinism and -optimism (Ke R błowski, Dobruszkes, and Boussauw 2022).Second, they tend to 'center upon on issues of utility, efficiency, and economic growth achieved through "rational" planning and decision-making' (Ke R błowski and Bassens 2017, 414), leading them to assume that commuter mobilities are linear, uncomplicated and bound to specific spatialities and temporalities, and thus to frame the FLM temporally and spatially.Third, such visions are shaped around the dominant figure of the commuter.They imply that commuters are able-bodied, independent, digitally connected and affluent, and correspond with the notion of the young middle-class professional male; a rational actor who responds to industry and policy stimuli as intended (Strengers 2013).
The most prominent FLM solutions hence tend to be designed as individual, service-based, and on-demand, often including AI-powered automated decision-making functionalities (ridehailing or integrated digital mobility service platforms).These FLM visions therefore frequently fail to account for those mobilities that are the most challenging or disregard the purpose and meanings of everyday mobilities which are inherently social in nature.Moreover, in the narratives supporting these types of visions, it is unclear what the first/last mile is and implies, both spatially and temporally.
We argue that to achieve the objective of enabling people to minimise individual car use, more attention needs to be paid to diverse mobilities and to the realities and contingencies of the FLM.We therefore aim to provide an understanding of the beginnings and endings of everyday journeys beyond their status as a societal problem related to unsustainable mobility patterns that can be solved by introducing automated technological solutions that, supposedly, make travel time-efficient and seamless through the physical space surrounding their homes and destinations.Instead, we analyse FLM mobilities on the terms of the localities of people's homes and destinations, which are complex lived spaces where ways of inhabiting and belonging are negotiated.As we show below, these mobilities are inseparable from wider social and material networks implicated in everyday mobilities.
Therefore, we interrogate and re-frame the 'first and last mile' concept, by questioning the pertinence of the spatial and temporal category of a FLM.In doing so we take a design anthropological approach which theorises everyday spaces, temporalities and practices as always emergent and contingent, and attends to the ways in which everyday life situations and possible futures are constituted and shaped through people's everyday experiences, practices and imaginaries (Pink and Salazar 2017).
Applied to future automated mobilities, design anthropology has already highlighted how human creativity and improvisation constitutes socio-technical relations and configurations that are unanticipated by dominant mobilities narratives (Pink, Fors, and Gl€ oss 2019).Our approach also accounts for how people continuously learn and adapt their mobilities as they encounter the complex social, material and technological environments in which they live.A design anthropological approach (Fors and Pink 2017) therefore highlights how people accrue mobility competencies in locally situated and changing contexts, which shape and are shaped by the social and material environments in which they live.
Here, we apply this lens to ask how people actually go about solving their everyday logistics in ways that are adaptive to the contingencies that arise in their in daily routines, relationships and environments as they travel locally.We analyse the particular situated mobility competencies, values and socialities that are constructed and enacted as people do so.In doing so, our ambition with this article is to contribute to a social understanding of what happens within the immediate space around the home and destination and the implication of this for how to grapple with 'the first and last mile challenge'.
To demonstrate this, we draw on ethnographic fieldwork into the spatial and temporal circumstances of the beginnings and ends of people's journeys in a peri-urban neighbourhood in Sweden.The insights from our analysis of these everyday mobilities and competencies contests and questions the assumptions underpinning dominant technological solutionist understandings of both the FLM themselves and the supposed challenges they present.
In the following sections we first discuss the most prominent discourses of FLM 'challenges' and their association with automated technologies and outline our methods.We then discuss three related dimensions of our ethnographic findings.As is conventional in design anthropological analysis, we use the dialogue between the findings and the insights they produce (rather than separating them): to question the direct causal relationship between automobility/private car ownership and FLM problems; to challenge the FLM as a spatial and temporal category of daily mobility; and, in doing so, to demonstrate how situated mobility competencies, values and socialities participate in generating local and workable mobility spatialities and temporalities.To conclude, we sum up and call for a reconsideration of the automated FLM and the benefits it promises, through attention to the diverse mobilities that already constitute the space-time it represents.

Context: automation and the 'last mile challenge'
Autonomous vehicles (AVs) and 'shared' Mobility as a Service (MaaS) or Mobility on Demand (MoD) systems have frequently been proposed as 'solutions' to a particular version of the 'first and last mile problem' applied to lower density or peripheral suburban areas (Legacy et al. 2019;Maginn, Burton, and Legacy 2018;Ohnemus and Perl 2016;Shaheen and Chan 2016).Effectively, this means that universalised automated technology solutions are being proposed (yet not successfully implemented), on the basis of unsubstantiated assumptions about how people engage with the everyday spatialities and temporalities that characterise the sites of departure and destination in their travel.In this section, we first outline current approaches to the FLM as a 'challenge' to be solved, before identifying a series of critical issues concerning this dominant conceptualisation.
Solving the FLM challenge is frequently presented as a key factor in encouraging transit use and in compensating gaps in transit coverage (Gurumurthy, Kockelman, and Zuniga-Garcia 2020).From a technologically-driven perspective, it is believed that poor connections from public transport nodes to people's homes are the main reason for people's preference of the privately owned automobile (Lesh 2013;Shaheen and Chan 2016;Mohiuddin 2021).This vision of FLM passenger mobilities has also attracted commercial interest as large transport companies seek to position themselves as multimodal door-to-door service providers (Bonneville 2018).In turn, car manufacturers and other industry actors have started investing in MaaS and MoD platforms, 'shared' mobility, automated vehicles, and micromobility technologies.
Shared mobility solutions have also increasingly been proposed as a way to 'address first-and last-mile connectivity with public transit, extend the catchment area of public transportation', and encourage 'multimodality for first-and last-mile trips rather than driving alone'.On-demand shared mobility services are regarded as a more flexible alternative to 'costly feeder bus services and land-intensive parking infrastructure' (Shaheen and Chan 2016).
While these solutions include less investment-intensive and more time-proven practices such as peer-to-peer ridesharing, a particular focus has been placed on automated and autonomous vehicles (AVs). 1 Autonomous 'pods', self-driving shuttles and personal rapid transit vehicles in particular have been promoted as FLM solutions (Jones et al. 2023).Beyond aspiring to provide access for elderly, low-income, and non-drivers, shared autonomous vehicles (SAV) are said to ensure the FLM of trips in low-density areas by integrating with public transport (Ohnemus and Perl 2016).Overall, despite calls to consider micromobility alternatives such as cycling (Behrendt 2020), debates involving the FLM have been largely focused on motorized, autonomous and networked vehicles and cars in particular.
Technologically-driven FLM 'solutions' represent a form of 'technological solutionism' (Morozov 2013), which assumes that the societal problem is that people use private cars to drive the FLM because they have not been provided with more sustainable, efficient alternatives.Such approaches see travellers as rational actors, failing to account for the extensive evidence of the importance of sensory (Kent 2015), symbolic and socio-technical (Sheller and Urry 2016) dimensions of automobility.These solutions also fail to account for the diverse ways in which people really travel their FLMs, or what these journey segments mean to them.For example, that passengers will prioritise elements of convenience and combined activities over simple time efficiency and modal choices; preferring longer transits with fewer changes in order to allow for comfort or in-transit activities.
Moreover, while the FLM 'challenge' is frequently presented as a homogeneous and technical issue, research has shown that the experience, practice and narratives of these journey segments varies widely across and within different cultures, contexts and spaces (Hickman and Vecia 2016).Thus, by overemphasizing the functional importance of the FLM (for modal choice), these approaches tend to underestimate its social, experiential and symbolic significance.Indeed, the concept of the FLM is closely associated with dominant figure of the mobile subject, most notably affluent and able-bodied male white-collar commuters executing regular to-and from work travels.However, daily mobilities are not limited to linear commuter journeys and are at least partly associated with everyday care, the responsibility for which tends to be unequally distributed, while mobility temporalities and spaces are gendered, and vehicles intended for emancipation may become vehicles for the extension of domestic work (Demoli and Gilow 2019).
It is therefore crucial that we provide locally grounded, in-depth understandings of traveller practices, experiences and (shared) meanings.To do so, we begin by critically unpacking the concept of the FLM as a spatial and temporal category and account for how such mobilities are produced 'close to home'.

Methods: situating the last mile through design ethnography
To access local meanings and practices of mobility we developed design ethnographic (Pink et al. 2022) explorations through our engagement with a suburban area in Sweden.Design ethnography involves an approach to research which, amongst other things, brings dominant technologically solutionist narratives into relief with the ongoingly emergent realities, contingencies, ways of knowing and learning and creativity of everyday life (Pink et al. 2022).Rather than working towards automation as a script for the future, our methods were designed to interrogate the possibilities of automation; to expose the power relations it implies and surface existing and desired everyday mobilities and future mobility transitions.Our methodology was designed to foreground the experimental and experiential characteristics of people's everyday mobilities and make sense of the messiness of everyday logistics (Fors et al. 2022).
We engaged this approach to reflect on the FLM concept in relation to real everyday life experience, by investigating how people living in a specific locality navigated, experienced and constituted the spatialities and temporalities associated in dominant narratives with the FLM .Such suburban spaces offer laboratories for sustainable change (Loubi e 2019), due to their complex environments, low public transport offerings, and frequently monomodal, car-centred mobility.Moreover, such peripheral spaces are undergoing profound changes (in terms of demographic composition, spatial organisation and built environment) that question the meaning and implications of suburban and peripheral spaces themselves.For instance, in processes of rural gentrification and continued suburban growth, agricultural spaces are becoming commuter territory, while simultaneously rural territory is being reinvested not as a periphery to the city, but also as place of activity, source of revenue and basis for more sustainable lifestyle.
Our fieldwork was undertaken in the suburbs of one of Sweden's larger cities, with a growing population of just under 7000 in a hilly semi-rural area stretching across around 70 km 2 , where distant clusters of residential housing are connected to each other and the city by a single fast road.The specific topography, infrastructure and public transport availability make this a particularly compelling example in relation to mobility transitions.
In this rural areawhich we call Monsite -20 km from the nearest city centre, where former summer houses had previously been transformed into permanent residences, new, dense housing development is challenging the existing infrastructure and demographic makeup of the area.Hedlund (2016) would term a 'middle-class countryside within the urban shadow'.Its population has high educational levels, most work in qualified sectors and form a significant part of the working population commuting to the urban centre.Simultaneously, residents and local organisations are turning their attention to local activities alongside the development of small-scale organic farming, adventure tourism and local food production.Lined on two sides by a nature reserve, lakes and wooded hills, the remaining fields and meadows are punctuated with horse stables, persisting farms and more isolated housing.

Monsite is part of what
Twenty-two residents were recruited using a snowball method starting from key informants from local neighbourhood organisations (a neighbourhood development committee, school, church and local sports clubs), and recruitment and outreach events held outside the local supermarket.Participants included 11 women and 11 men, aged between 14 and 77, living in variable household compositions, and 11 were parents of school-aged children.The sample was designed to reflect the importance of family logistics in local mobility and we interviewed several members of the same family (partners and/or their children) in three cases.Participants lived in different neighbourhood types, from dense housing clusters to sparsely populated areas, from remote locations to accessible neighbourhoods closer to the fast road system.
To investigate how people experience and practice FLM mobilities under the restrictions of the COVID-19 pandemic, we combined individual online interviews with on-site visual ethnography.In our online semi-structured interviews of around an hour each, we focused on participants' biographical narratives, residential trajectories, perceptions of their neighbourhood and existing mobility and sharing practices that inform their everyday mobilities.We also enquired about potential improvements and future developments participants envisioned for their immediate environment and mobilities.This provided insight into participants' daily mobilities, their motivations and social context for modal choice as well as their representations of different forms of (future) mobility.
Prompting participants with the question of automation moreover provided the opportunity to confront diverging narratives of future mobilities, thereby offering a template to emerge frictions in everyday mobilities and proximate spaces.These video-recorded interviews were supported with shared maps that participants could draw on, which allowed us to situate mobility practices, meanings, and projections within the geographic space and to gain insights into routine routes and meaningful connections between different places.
These elements were further investigated through an innovative research method we developed to explore participants' practices in more depth: the two-car convoy (Brodersen 2021).Instead of 'in-car video ethnography' (Pink, Fors, and Gl€ oss 2019), or fixed video-cameras installed inside cars (Laurier 2013), our method involved participants driving their own cars to guide two researchers who followed them in a second car.
Participants chose the starting point and determined the routes in relation to relevant places and roads identified through a set of initial questions.While driving, participants and researchers communicated via mobile phone and the encounter was video-and audio recorded.In this neighbourhood, which is too dispersed to be walkable and where car travel is the dominant practice, this technique encouraged participants to identify the spaces most relevant to (their) mobility, and to explicitly share their situated practices and embodied knowledge.This innovative combination of traditional interviews, online, and mobile, go-along methods (see Merriman 2014) allowed us to learn about the layout and meanings of local spaces and how existing and imagined mobility decision-making practices were embedded in the socio-spatial context.

Living the 'first and last mile'
We now draw on our ethnographic findings to demonstrate three key points.First, by showing how participants solve their everyday mobilities and examining the role played by automobility in these practices, we complicate the notion that car travel and ownership compensate for public transport accessibility in the FLM.Second, we question the FLM as a unit of space and, third, as a unit of time, by exploring how participants' mobilities unfold at the beginning and end of their everyday journeys and around their homes.

Living in Monsitecomplicating automobilities
Most of the participating households in Monsite rely on two or more carsprivately owned or company cars, often a combination of bothto ensure their everyday mobilities.In their neighbourhoods, using several cars per household forat first glancemonomodal car travel, appears to be the norm.While this is in part explained by ease, flexibility and greater comfort (see also Kent 2015), the motivations, rationalities and actual practices of automobility were more complex.
In participants' narratives, family logistics are key in justifying car use and ownership.Gemma (49), for example, who lives in the area with her husband and two pre-adolescent children, has experimented with having one car or no car while living in the area.Having a connection to the public transport system was initially a key factor in choosing their house when they moved to the area several years ago.They are reluctantly leasing a second vehicle in addition to her company car to help solve their 'puzzle' of everyday family logistics, but Gemma and her husband would much rather avoid the complications and responsibilities that come with car ownership altogether.As for Gemma, who went through stages of having no car to needing several, a recurring justification for multiple car ownership is the need to drive children and other dependents to different destinations, particularly extracurricular activities inside and outside the area.
However, trips made entirely with private cars often combine diverse activities and journeys.Some of these combined mobilities serve as an argument for the car as a single transport mode by many participants.Simon's mother (47) uses the car to drop off her son at school on her way to taking the dog to day care before work, and Felix (44) stops at the small supermarket on the main road on his way back from work.Gemma picks up parcels at the local gas station on the way to the football field or when she does the shopping on her way home from work.This 'trip chaining' (Scheiner and Holz-Rau 2017) as a strategy to handle the complexities of family logistics and care work also justifies the use of the car.Other than a 'mobility of care' (S anchez de Madariaga 2016), these combined trips join various layers of activity, social spheres and uses of local space.Travelling by car is a way not only of adjusting to the demands of different household members but also to account for the serendipity of moving through everyday life spaces.
Even households with several cars routinely combine different modes of transport in creative and adaptive ways to 'solve the everyday puzzle' and to cope with the changing challenges of daily routines, seasons and environments.Residents use commuter parking spaces, but also more informal and less infrastructure-driven hubs for their coordination.Antonia (42), who lives further from the main road, drives her kids to school and drops off their bikes at the nearest bus stop on the way on days where they finish early, so they can take a tram and bus back and cycle the last 3 km home along a dirt road.In the winter, she prefers to coordinate with her husband to either pick up the kids at school or at the bus stop.These specific combinations of multimodality, coordination and adaptation serve to fit various objectives and constraints and can be part of collaboration within families and communities.Amanda (42) drops off her daughter's team-mates at the bus stop after practice so they can walk or bike home from there, but when it is dark, she tends to drive them all the way to their door.Olaf (45) and a few of his neighbours use a post-box at the bottom of his street as a collection point for children to drive to football practice together.Felix used to drop off his sons at day-care and the bus stop respectively on his way to work, then his older son would pick up his little brother when walking home from the bus stop after school.Additionally, participants also realised a number of their journeys without using their car, sometimes combining several other modes of transportation.
As these examples suggest, the linear journey modelled on home-to-work commuter travel does not reflect the way (auto)mobility was understood, experienced or performed by participants in our research.Their journeys were overwhelmingly combined activities that were not limited to a single point to point journey that would take them outside the area.And while difficulties in reaching the nearest public transport stop were mentioned as part of the motivation for using the car, the journeys into which participants invested the most effort and planning were multiple short drives within the area or surrounding boroughs.Below, we show that the actual characteristics of this space within the first few kilometres around the home, its material qualities and meanings, influence how mobilities are organised and conceived of and complicates the notion of the FLM itself.

The spatiality of the first and last mile
What space counts as the FLM and what do we really find there?As noted earlier, recent uses of the FLM concept referring to commuter travel have defined this space as representing a deficit where transit gaps are found at the beginning and end of travel.In Monsite, this space can surpass the literal mile or kilometre; although the area is connected to the urban public transport system via an express bus route, the nearest bus stop can be more than 5 km away.The route from the main road is often up a steep uphill and winding road with little pedestrian space and low visibility.Navigating some of these sparingly lit and often unpaved streets requires intimate locally-accrued embodied knowledge, particularly in winter.This recalcitrant space makes common micro-mobilities challenging and often unlikely.Hence, this space exceeds the last mile or kilometre in scope and its particular qualities resist the most commonly proposed FLM 'solutions' while at the same time having considerable effects on everyday mobility practices.
Single mode transport or multi-car ownership are in fact often motivated by both lack of infrastructure and constraints posed by the built environment.For instance, Pernilla (42) lives along a busy road about 4 km from the nearest school that she deems too dangerous for her children to bike or walk along by themselves, given that there is no cycle or pedestrian path, little to no lighting and utility vehicles driving at high speeds.This means that she is unable to take the bike and bus to work, as she would when her children had access to a school taxi, and unable to offer her children the desired autonomy: … if I could take the bicycle every day and the children could go … on their own to school, maybe we could have only one car.If … we had a bicycle path all the way to Torstorp and to Bigsby I could say we could manage with one car.
While a lack of transit options is used to explain dependency on cars, it is, however, often the commuter-oriented linearity of transit options which is misaligned with everyday life.In contrast, in Monsite social relations and trips serving the domestic sphere extend into the surrounding boroughs more often than towards the city: 'You can't really go with the tram horizontally or across, you can go towards the city centre' (Amanda).Thus, it is sometimes the structure of the organisation of travel, rather than the actual distance to travel hubs that is decisive for participants.
The immediate radius around participants' places of residence (be it a mile or 10 km) frequently concentrates numerous interconnected short distance travels to key destinations; it is also dense with social relations, activities, and meanings.In some close-knit housing clusters, children can roam more freely, parents give accounts of their ideal image of children knocking on each other's' doors to play.The immediate environment will contain 'favourite places' which are not always the most frequent destinations but carry special value.
Rather than wishing for automated modes of leaving the locality, Monsite participants highlighted the need to improve the quality of local space, to make many trips redundant, improve safety and allow for more meaningful interactions; including needing more schools, shops and social spaces.As she guides us through the area, Lisa insists on showing us the places that function as central travel hubs in the area, which she perceives as 'boring' and lifeless, to demonstrate why she chooses to concentrate her activities elsewhere.Although local mobilitynot the commuteis usually portrayed as being the most problematic and automobile by participants, some like Mats and Linda stress the strain of commuting.However, they would rather develop their life locally than optimising their commute: Linda: 'We don't want to travel back and forward to [work].
Mats: We want to stay here.And work here; we are fed up with traffic and the stress in the morning, in the morning traffic it wears you down'.
Improving local life thus appears as a higher priority than improving transport options for many of those who already have the resources to ensure and maintain their mobility despite difficulties (Ollivro 2005) and critical for those who do not.

The temporality of the first and last mile
In this section we critically examine the temporality of the FLM, at the outset and end of everyday journeys, to show what participants do within these times, and how they define and delimit them.For instance, the time that begins and concludes everyday journeys can be perceived as valuable time which, other than the cost of getting from A to B, is invested with multiple activities and serves a number of other functions.Amanda (42) works as a teacher in a school not far from the city centre.She lives with her husband, three adolescent children and an infant in a newly built house on top of a hill.When she travels to work, she takes the express bus to the urban hub closest to her work and then, instead of taking the tram that would take her almost directly to her destination, walks the last stretch up a slight slope towards school, repeating the same on her way back.To her, this time works as a decompression chamber, as a space in which to transition between social spheres and times (see also Hubert et al. 2005), between private and professional life.It enables her to fit exercise into her day and to wake up and wind down, listen to podcasts, reflect and anticipate.It is also the only time Amanda is alone.In contrast, Simon (13) explains during his interview how he coordinates with his friends via text message to make sure they are on the same bus so they can spend more time together.
Rather than optimising the speed and efficiency with which she reaches her destination, Amanda intentionally builds a time-space into her day that is valuable in itself.This resonates with previous literature on the use of mobile times (Bissell 2010;Clayton, Jain, and Parkhurst 2016), and qualitative dimensions of mobility determining the preference of routes and modal choice rather than efficiency.This questions the idea of the FLM as a 'challenge' to overcome and repositions the beginnings and endings of journeys as part of a larger time ecology.
Amanda's ritual also reflects the value attributed to active mobilities.Health and the joy of physical exercise influence the perception of modal choice, routes and the experience of everyday mobilities, often more so than considerations of efficiency and optimisation.Encouraging exercise and autonomy for children becomes an essential part of parenting and is mentioned as a reason for favouring cycling and walking, for expressing guilt or regret over single mode car travel but also as a reason to express doubts about the use of automated, motorised, on demand mobility services.During interviews, participants conceded that access to an automated vehicle or service for their children might carry advantages in terms of household efficiency but did not necessarily fulfill their desires for their children's mobilities.Felix insists that rather than having an automated on-demand service take his children wherever they need to go, he 'would prefer that my kids are walking, that they are moving and they take the bike [ … ]'.
Not only do these types of mobilities carry valuethey also question the idea of the 'last mile' as travel time and effort to be eliminated.Felix's concerns about his children's activity also reflect a key dimension of modal choice and vehicle ownership: the way parenting is conceived of in relation to mobility.In fact, although ferrying children to activities is mentioned as one of the main reasons for multiple car ownership in interviews, and participants claim that being relieved of this duty would be a priority for them to improve their daily mobility routines, driving children is an elemental part of parenting in Monsite.While there are diverging perceptions about whether the time spent together in the car can be defined as 'quality time', providing mobility is a way of participating not only in children's lives but also in the local community.
In Monsite, this is linked to the way life cycles and residential trajectories are connected: as people move in from elsewhere to acquire property, children are a key factor of integration and people with very young or adult children sense the difference.Hans, whose children are grown, notes he is longer as integrated in the neighbourhood as he was, while Paul, whose son is not yet in pre-school, says he has little to no activities in the area.
As part of local sociality (Pink 2008), driving children or otherwise dependent people is also a key site where practices of sharing are built, negotiated and considered.While 'sharing' services are frequently proposed as 'solutions' for closing gaps in transport systems (Schaller 2021), especially in relation with SAVs (Ohnemus and Perl 2016;Moorthy et al. 2017;Huang et al. 2021), varied and complex sharing practices are already established within the area, and these are inextricably intertwined with relationships and identities within groups, organisations and family networks.
Although participants did mention the potential to expand some of these sharing practices, the group of people with whom participants are willing to share is not limitless and does not necessarily accommodate anonymity; open and commercial sharing networks are unlikely to replace appropriately the existing sharing practices that are embedded in meaningful relationships, which they help to consolidate.Like driving itself, sharing the driving task appears as a part of parenting, but it also functions as a way of performing parenthood that extends beyond 'cultures of mothering' (Dowling 2000) or even processes of 'doing family' or reproductive mobilities that have been previously identified (Waitt and Harada 2016).Beyond this performative quality of mobility (potential), owning a car means being able to participate in a common effort and, by extension, in relevant local groups.Amanda shares the driving to track and field practice with other parents on the team: We have a super text group -often the night before, I write "I will drive to the practise tomorrow, I will pick up at that time at petrol station" and then the other parents write "yeah he will go with you".There's not the problem that no one wants to drive it's the other way around, so it's a good group, everybody wants to help out.
For children and young adults themselves, being able to drive and having access to different vehicles is also a key part of their integration and the continued construction of their identity locally.The coordination of mobilities and access to vehicles is a key site of sharing within families and groups.Eva ( 16) borrows her father's moped since she has been allowed to drivethis not only spares her the difficult 10 km bike ride back home after a taxing sports practice, it also gives her the possibility to give her sister a lift, relieving her parents of a lot of driving duties.For Oskar ( 16), his moped is central to socializing 'on the move', since he is part of a group of a dozen teenagers cruising the area on motorized two-wheelers for fun.Thus, times around the end points of mobility become extended when integrated into wider sets of activities and relationships.When the act of driving equates with acts of caring, the functional efficiency of point-to-point mobility in the last stretch of a journey as a time period can become secondary to surrounding practices and combined mobilities.
The frictions associated with mobility also extend largely beyond the actual travel time.A substantial part of producing mobilities is the time spent preparing and coordinating these mobilities, not only in the context of shared driving, but as part of everyday routines.Coordination and anticipation are a large part of maintaining both individual and collective mobility potential, yet this effort is usually not equally distributed within and among households and is often gendered (Gilow 2020).Situating the everyday experience of mobilities at the individual level, as do the FLM concepts analysed above, locates mobility coordination primarily inside the household unit, which reproduces existing inequalities.It moreover obscures the limits of automation as a solution for locally situated mobilities because it does not acknowledge these inequalities.

Automating the first and last mile?
Our ethnographic research revealed how the temporal and spatial categories of the efficiencyoriented FLM are misaligned with socially organised locality-based mobilities.In this section we draw further on our ethnographic findings to show, moreover, how locality-based mobilities, which are contingent on and responsive to the complex needs of everyday life, resist simplified, universal automated FLM solutions.When prompted to imagine improvements to everyday mobilities and presented with the possibility of automation, our participants considered diverse possibilities, many of which did not involve automation, and pointed out the limitations of automation in existing mobility practices and spaces.As we explain, their experiences and imaginaries of mobility furthermore suggested that automation may be missing the point of future mobility issues.

Situated mobility and local ways of knowing
Our ethnography showed that rather than seeking to rationalise their FLM mobility, participants had acquired locally situated knowledge and mobility competencies.The experience and expertise associated with these competences subsequently guided participants' anticipations of automated mobility solutions.Two central dimensions of situated mobility that emerged from our ethnographic analysis are key for understanding the social dimensions of FLM mobility: first, mobility is continually learned and adapted in relation to locally specific materialities and relationships, as participants incrementally developed and coordinated shared local social routines and practices.
Second, the resulting mobility competence plays a key part in the human agency and autonomy that underpins locally situated mobilities.These mobilities are situated in that they are inseparable from local materialities, timelines, relationships and networks of heterogeneous actors.Mobility competence, then, is situated in the socio-material time-spaces they evolved in conjunction with and, as such, are not universally scalable and transferable.However, while imposing a limit to automation, this situatedness does not limit the competence itself.Rather, it is through this situated learning, knowing and sharing that mobilities are experienced as carrying sense, social meaning, and collective agency.Dominant visions of automated FLM mobilities are not aligned with these already-existing situated mobility competencies.
For example, the particularities of local space and its recalcitrant nature were cited as reasons why residents anticipated automation to be difficult and doubted the viability of a systematic, ubiquitous AV system: I just have a hard time seeing how self-driving cars would work in real life.I would want to know the technology behind, how it works, if unpredictable things happen around the car.And if you would go on a tiny road, like the last two kilometres to the lake where I like to go … it's looking out for animals, since it's in the forest.And then also driving up the steep hills with the tiny stones in the ground I need to make sure that I can drive up safely, and not having the car getting out of my control and sliding down the hill again.(Emma,20) Beyond simple practical challenges, the anticipated difficulties Emma mentions in implementing AVs in relation to the material specificities of their local environment stress the value placed on locally situated mobility competence and learned ways of inhabiting local space.During the two-car convoys, these elements were integral to how participants guided researchers through the area; while driving, they anticipated narrow spaces and gave instructions on how to let other drivers pass or a change in road surface or an upcoming interruption in mobile networks and they instructed us on how to interact with other drivers on roads too narrow for two cars to pass.
Unmapped roadblocks, shortcuts and alternative routes, gravel roads, potholes, sharp turns and abundant vegetation do not lend themselves particularly well to automation.At the same time, they confirm participants' perception that their area is neglected in terms of urban development and infrastructure.Appropriating the layout of local roads, developing knowledge about minute difficulties and informal solutions, navigating the routines and language of sharing these spaces with other users, is part of inhabiting the areaof belonging.It is also part of the specific agency associated with mobility.The value placed on this intimate and incorporated knowledge thus extends beyond more functional concepts such as 'motility' (Kaufmann, Bergman, and Joye 2004) and is not as such replaceable through the delegation of driving to automated technologies, services and systems.
Participants also placed significant value in autonomynot only the ability to access destinations and services but also a mobile coming of age and a performative quality of mobility potentials.Participants do not cite a lack of information about mobility options as the central concern in their everyday mobilities; on the contrary, displaying a mastery of the different mobility systems, the particularities of local space and their interconnections was a way for people to display appreciation and appropriation of their local environment.For young people, owning and controlling their own vehicles was itself valued.For instance, as we met Antonia for a drive-along several months after her online interview, her daughter had in the meantime, gained a licence to drive her father's moped.As she explained: The moped has been like a gift from heaven because it gives her so much independence.And she really likes it.It is also great for football because if you have two hours of really intense training it is kind of hard to jump on a bike and go 7km.
Even when prompted with the possibility of a 'shared' automated service that would provide the same level of access and flexibility, teenagers and young adults insisted on the importance of owning and driving a vehicle.The association of the car with freedom and autonomy for young people especially in intermediary and low-density spaces with limited transit coverage has been well established (Drevon, Ravalet, and Kaufmann 2019;Demoli and Lannoy 2019).
The last mile, then, is also the site of conflicting versions of autonomy.On the one hand, automation is poised to promise greater autonomy to those who are excluded from driving or unable to access relevant destinations and activities by their own means.On the other hand, embodied and shared knowledge and practice of locally specific space and the interpersonal negotiation of ownership and sharing are part of what is considered to be a mode of autonomy that is central to the agency of the mobile subject.

Situated mobility values and local socialities
In addition to the local material situatedness of everyday mobilities, participants' practices and perceptions of sharing mobilities were inextricable from the relationships, roles and groups that they contribute to creating and consolidatingfor instance, when driving and sharing the responsibility for driving become local practices of parenting, as described in the previous section.The situated knowledge and practices of inhabiting local space also had a social component as our examples of sharing mobility experiences demonstrate.Oskar (16), for example, provided a detailed map of the most fun roads to cruise along with his friends on their mopeds, pointing out long passes for acceleration, challenging bends to negotiate or empty sections to practice wheelies with his friends.
The frictions of the coordinating and planning such socialities (Pink 2008) require are not problems that could be solved by FLM applications proposed for automated technologies and services.Rather, practices and relationships of sociality and care shape people's everyday mobilities (see also Balcom Raleigh, Kirveennummi, and Sari Puustinen 2020) and as such fundamentally undermine the notion that a linear and automated FLM solution could deliver the social needs that they accommodate.Sociality, here, refers to the ways in which mobility functions not only as a way to connect spaces but also to produce a social dimension of 'local interconnectedness' (Pink 2008).A 'mid-range conceptualisation' (Amit 2015), sociality here helps to think the processual and contingent (see Long and Moore 2012) forms of belonging, relating, experience and engagement by which last mile mobilities produce collective agency and shared local ways of knowing.
The 'last mile' is an opportunity to be 'brought together' but it is also a social, rather than a spatio-temporal concept; it is socially produced and contingent upon continuous actualisation.This implies, first of all, that it is subject to change and, second, that it inherently resists the type of change through predictability that automation requires where, in order for innovation to be dynamic, its socio-spatial counterparts must be static.
Moreover, the values associated with the mobility space-times as constituted by participants meant that mobilities that might be framed as FLM mobilities were in fact experienced as personal time and moments of transition between social temporalities.Participants favoured the improvement of local services and meeting places over technology-based services to connect different modes of transportation, and they valued opportunities for exercise rather than striving towards greater comfort and efficiency.Coordinating everyday mobilities is part of how people perform meaningful social roles; it reinforces locally situated relationships, affiliations and meanings of place, but also frictions and inequalities.Hence, there are clear limits regarding the role automation could play in improving everyday mobilities that are inherently social.
These insights invite us to reconsider where the 'challenges' in local mobilities lie and suggest that seeking to optimise the time people spend in the 'last mile' of everyday travel individually is not necessarily the right solution.Apprehending the extended time-spaces of mobility as an always contingent and contested 'commons' (Nikolaeva et al. 2019), we instead propose to focus on the situated socialities and values that everyday mobilities rely on.
Prompting imaginaries about automated futures with participants was illustrative in that it surfaced the local frictions and the social and material relations that characterise existing mobilities.But rather than providing indications for the implementation of FLM automation it revealed the tensions that arise where there is misalignment between everyday life realities and dominant narratives, ideals and interests in automated mobility transformations.Hyperlocal infrastructure, contingencies in public service and the quality of local space all play major roles in determining modal choice and everyday logistics.
As we have demonstrated, much of what is most central to everyday mobilities cannot be substituted or reproduced with automation.Moreover, focusing on automation risks detracting attention from mobility related questions relevant to participants' situated mobility competencies, values and socialities, and to the actual time-spaces they travel in and through.

Conclusion
In this article we have suggested re-thinking the FLM 'challenge' through the lens of the situated everyday socialities, competencies and materialities of close-to-home mobilities.The idea of automating the FLM implies a challenge which requires a technical (rather than social) solution, and it assumes mobility is an individual activity rather than collective, shared and continually negotiated.Instead, we emphasised the need for attention to: what everyday transport means to people; the social learning and coordination it entails; the particular terrains and infrastructures it navigates; and people's expectations of how automated mobility solutions should support their everyday logistics.
Our findings both question the universal applicability of automated and standardised mobility 'solutions' and the central place given to technology-driven, individualised versions of automation in future visions of FLM travel, because they fail to account for the most salient factors in participants' discourses around everyday logistics and their modal choices.Understanding automation as a generalised solution and the last mile as a homogenous problem distracts attention from the factors that support participants' situated production of mobility and could at worst result in adding another layer of imperfect infrastructure to an already existing mesh of flawed physical infrastructure and mobility services.
In a context where many of the costs and challenges of mobility are not just related to commuting but to the task of realizing everyday household logistics, we have shown that mobility extends beyond the act of driving and encompasses a range of sharing and coordination efforts.These mobility practices are entrenched with care work in various types of relationship and more generally serve as a vector of social integration (while also reproducing existing inequalities).Our fieldwork hence serves as a reminder that improving mobility near home is not simply a matter of improving linear access to public transport options.Automation might relieve people from driving, however, it does not replace the act of caring, risk mitigation and social integration that is implied in the production of mobilities.
Moreover, dominant images of automated mobilities frequently focus on the individual commuter, travelling in a controlled inside environment, sheltered from their surroundings (Manderscheid 2014).These imaginaries reproduce dominant images of mobility and commuters and represent gendered and class-based mobility norms.They are informed by the notion that the benefits of automation are self-evident and that the barriers to these benefits lay in securing its adoption and the trust of potential users.These understandings are equally written into dominant narratives of the FLM which seek to provide technical solutions that have the potential to modify people's behaviour.Posing the last mile as a central issue demanding a technological solution reiterates these exclusive images of mobility.It moreover misses the point of where the frictionand future imaginationin everyday mobility lies.
To understand the dynamics of the FLMs of any journey requires acknowledging the complexity of mobilities.It involves analysing how different mobilities connect to each other and to other elements of the material and social environment, rather than focusing on the connectivity of specific geographic locations to a particular transport network or infrastructure.Techno-solutionist responses to the so-called FLM challenge, where the technologies would solve transport problems for people within the mile around their homes, turn our attention away from how people already manage their everyday mobility logistics and contingencies.
Going forward, we therefore suggest a research agenda focused on emerging practices and future imaginaries that accounts for the situated, social and uncomfortable complexities of everyday mobilities and emancipate themselves from technology-driven universalist scalabilities.This focus on the social and material time-spaces constituted by everyday mobility we propose does not solve the FLM question.Instead, it offers a new starting point from which to consider how automated technologies could enable already existing socio-spatial infrastructures to develop in sustainable directions.
Note 1 'Automated' and 'autonomous' vehicles are being used almost synonymously throughout large parts of the literature irrespective of the fact that even with high levels of automation, most vehicles and mobility systems are not strictly autonomous from human control.In this paper, we focus on the wider notion of automation as general principle for addressing mobility objectives beyond AVs (Jones et al. 2023).