Community Smarts in Grassroots Initiatives to Support Cycling and Walking in Large Urban Areas

ABSTRACT In the popular imagination, smart mobility, much like the smart city, is a high-tech utopia of mobility efficiency and environmental sustainability. However, a growing body of literature is questioning the promises associated with “digital solutions,” particularly with respect to the neoliberal rationales behind their deployment. On the other hand, relatively few scholars have taken the same critical analysis to smart mobility. In this article we draw on interviews with leaders of grassroots organizations that support walking and cycling among marginalized communities in São Paulo and London to argue that “smart” interventions in active transportation do not have to be top-down, technocratic, or necessarily digital, though they may be so, when informed by community input. We suggest that the locally sensitive knowledge behind grassroots initiatives may demonstrate a form of “community smarts” that enhances capabilities and itself informs whether or not, and if so how, to use digital technologies to advance social and environmental missions.


Introduction
A large multidisciplinary body of literature has asserted the centrality of "smart" technologies in efforts to bring about more sustainable societies.In transportation, for instance, smart digital technologies, from mobility-as-a-service apps to sensors embedded in smartcards, bikeshare systems, and connected and autonomous vehicles, are widely believed to facilitate shifts to cleaner, healthier, safer, and more resilient mobility systems.In popular discourses ranging from the mass media to government communications and even the campaigns of some large cycling organizations, the enactment of "smart" cycling tends to be represented as contingent upon high-tech digital technologies, top-down innovation models, and willing governments.However, the number of voices asking what exactly "smart" means, and in what ways a smarter city, transportation system, and even velomobility will be not only greener but more socially just, are growing (e.g., Luque-Ayala and Marvin, 2015;Lyons, 2018;Nikolaeva et al., 2019).
In this article we aim to contribute to thinking along this line by analyzing the role of digital technologies for grassroots initiatives that support cycling and walking in the cities of London and Sao Paulo, and by considering how the adjective "smart" can be reclaimed in a manner that is appropriate to the particular logics and modes of operation of such initiatives.We propose the notion of "community smarts" to connote the collective and interactive forms of intelligence that animate, and are cultivated by, grassroots initiatives supporting cycling and walking.
In this context the term "grassroots initiative" refers to the actions, projects, and/or schemes of activists and civic society organizations that seek to realize an ambition or goal, such as making more people cycle or walk, in a bottom-up manner.It is a generalization of grassroots innovation, which in sustainability science literature is used to denote the development of: … novel bottom-up solutions for sustainable development … that respond to the local situation and the interests and values of the communities involved [and] share [a] commitment on the part of those involved towards openness and inclusion in the processes of innovation and the outputs of innovation.(Smith and Seyfang, 2013: 827) The polysemous term "smart" has a long history with uncertain etymological origins.For the past 500 years the adjectival use of "smart" has inferred speed, neatness, capability, impudence, stylishness, and, of most relevance to the present topic, cleverness, intelligence or knowledgeableness, in reference to people, actions and/or remarks (OED, 2021).From the mid-nineteenth century, the term has also increasingly been used to describe machines or objects that are, "designed to act or respond to conditions in a more sophisticated way than is typical," and that, "specifically [contain] a microprocessor (opposed to dumb [sic] …)" (OED, 2021).The OED's antonym is significant: anything that is "not smart" may be said to be "stupid."The use of "community smarts" may be seen as redundant-few would question that a community possesses some sort of adaptive intelligence-but, based on our empirical case, we propose the term here as a counterpoint to the connotations of the compound term "smart mobility" that markets digital technologies as inherently intelligent and intrinsically appropriate enhancements of transportation systems.
We argue that attempts to make transportation more sustainable do not have to be top-down, technocratic, and reliant upon digital technologies to be "smart," though they may sometimes be so.They also come in bottom-up smart modalities, which are typically enacted in grassroots initiatives.In these modalities both the overarching objectives and the role of digital technologies differ from what is conventionally understood as smart mobility.The focus is not on maximizing profit, increasing transportation system efficiency, or encouraging economic growth, but instead on increasing social equity and justice, community cohesion, and reducing transportation's environmental burden.A key concern is often to enhance people's capabilities (Sen, 1999) to move in environmentally and socially sustainable ways, particularly among those who suffer 'transport poverty' (Lucas et al., 2016) because they lack access to forms of transportation, experience difficulties in reaching meaningful activities at destinations, struggle to afford transportation, and/or are disproportionally exposed to risks associated with transportation (e.g., traffic injury, sexual harassment, exposure to air or noise pollution).In grassroots initiatives digital technologies are thus neither meant to increase the "efficiency" of urban systemic circulation, nor deployed specifically to amplify the general uptake of cycling or walking throughout a city-objectives that tend to be at the heart of corporations' and entrepreneurial governments' smart transportation visions.Instead, the relationships with digital technologies such as sensors, social media, and geovisualization tools are contingent, pragmatic, and tactical: those technologies are used when convenient, appropriate for the specific and often changing purpose at hand, and offering a frugal option, but abandoned or disregarded almost just as easily.They are just one of many tools that can be used to enhance mobility capabilities and reduce disadvantage in particular parts of a city.This scalar dimension is important: because most cycling and walking-related grassroots initiatives are small in terms of spatial reach, staff and volunteers involved, and available financial budgets; they are often tailored to the needs of specific groups, communities, and localities within a city, rather than standardized across cities.In this context "smartness" is dependent not so much on the "bird's-eye" perspective (Evans, 2011 ) afforded by masses of data assembled with the help of extensive and distributed sensor networks, but on "sensible" 1 use of digital technologies alongside, and informed by, the embodied yet distributed knowledge and learning that is developed from "dwelling"-the repeated, practical, and haptic immersion (McFarlane, 2011) in interactions with the community or communities a certain grassroots initiative seeks to support.
To broaden academic understanding of smart mobility and elaborate the notion of community smarts, this article first reviews some of the critical literature on smart cities and mobilities, and then draws on interpretive analyses of transcribed interviews with organizations behind grassroots initiatives supporting cycling and walking that took place in London, UK, and São Paulo, Brazil.The empirical reportage is organized around four themes: the selective use of social media; seeing beyond numbers; the limits of geospatial technologies; and ethical and pragmatic tensions between local communities and extractive global IT companies with respect to the sources, ownership, and uses of geographic transportation knowledge.

Reimagining Smart
Conceptions of "Smart" in the Smart City, Smart Mobility, and Smart Cycling Discourses The term "smart" pervades contemporary discourse on mobilities and the city.Cursory Google Scholar searches for "smart mobility," "smart transport," and "smart transportation" in the title of journal articles yield 1,720, 335, and 966 results, respectively (date: July 17, 2023), each having doubled in the past three years.Specialized academic journals (e.g., Smart Cities, IET Smart Cities) have been well-established, and there are multiple MSc programs focused on smart cities at leading universities.Even by 2014 there may have been as many as 23 distinct definitions of the smart city circulating across a range of academic literatures (Albino et al., 2015), many of them promising grand futures and smart utopias (Luque-Ayala and Marvin, 2015;Marvin et al., 2015).Some commentators have gone so far as to claim that "it won't be the minds of humans, but those of machines, that will most fully understand the world" (Rees, 2015:11).The gray literature is even more likely to associate the "smart city" with digital technology.Consider, for instance, the UK's Parliamentary Office of Science and Technology recently declaring in a research briefing that the term "'smart cities' describes places that incorporate a range of technologies (especially those that collect and use data) to address economic, social, and environmental challenges" (UK POST, 2021).
As far as the meanings of "smart" in relation to the city or transportation, scholars have lamented the dearth of discussion (Garau et al., 2016;Luque-Ayala and Marvin, 2015;Lyons, 2018;Schwanen, 2018).The term smart itself is a precarious moniker, with some suggesting its interchangeability with synonyms (e.g., intelligent) while others define the term as distinct (Gaffney and Robertson, 2018: 48).While the term smart cities was initially associated with smart growth and New Urbanism, it soon became dominated by reliance on digital technologies (Gaffney and Robertson, 2018: 48).Nonetheless, the smart city literature has now arguably bifurcated into separate technological versus social streams (Joss et al., 2019).
It has been argued that people use "passive, 'dumb' technologies, like shovels, but interact with smart technologies that exist to help us, serve us, to make our lives easier, and more interesting" (Brenner, 2007:4&5).But this view is problematic in two related ways.First, like much of the smart city/mobilities discourse, it may anthropomorphize technology (Lyons, 2018) by inferring dialogue.Similarly, it considers only the discursive as capable of interaction and feedback; for instance, others have argued that as mediators between the human and the world, even simple technologies, such as the walking stick of a vision impaired person, may provide such sensitive feedback that the distinction between the technology and the person becomes ambiguous in its hybridity (e.g., Michael, 2000;Nixon, 2012).This article takes the position that regardless of feedback mechanisms and algorithms, people still use digital technologies as tools (whether or not the tool is best suited to resolving a particular problem), and may also interact with other human beings.
The term "smart mobility" has generally superseded the earlier term "intelligent transport," possibly because of ICT (information and communication technology) industry marketers' recognition of the friendlier feelings associated with the former compared to the elitist connotations of the latter (Albino et al., 2015).In the mainstream smart mobility literature, the term, "refers to the provision of mobility options and services that are purportedly efficient, green, and inclusive, and that usually rely on the operation of digital technologies" (Schwanen, 2018: 65).They are also "assumed to be an axiomatically positive manifestation of technologically-based development in transportation systems, services, and their use" (Lyons, 2018: 6).Common to many of these accounts is the seeming inevitability of (future) mobility's dependence on digital technologies.While a few authors have actively challenged this dependence (see, for instance, Popan [2019] on cycling), the more common view is that smart technologies will radically alter current mobility systems and practices (see, for instance, Canzler and Knie [2016] on automobility and Behrendt [2020] on velomobility).It could be argued that the majoritarian perspective too easily forgets that "data and the digital are the product of human interventions and decisions" (Wiig and Wyly, 2016: 489, emphasis added).
When it comes to cycling specifically, it would seem that "smart" is increasingly understood in terms of connectivity, (big) data, and the Internet of Things.Behrendt (2016Behrendt ( , 2020)), for instance, conceptualizes smart velomobility in terms of networked and power-laden technologies, practices, experiences, infrastructures, and broader systems around cycling that transcend the individual/collective, physical/digital, and human/machine binaries.In this way the common perception of cycling as a sustainable but offline and low-tech mode of mobility is challenged.Nikolaeva et al. (2019) build on this conceptualization, drawing out a range of connotations of smart cycling from mainly Dutch website texts that include but are not limited to connectedness, a shift from ownership to use of bikes (e.g., sharing/mobility as a service), customization of experience, digital surveillance, and elegance and fashion.

Critical Thoughts on Smart Mobility
Compared to the critical scholarship on smart cities, critical work on smart mobility is more limited in size and orientation.The former has unpacked, through a wide variety of case studies, the way alluring narratives of "smartness" are composed by the private sector as a post-political techno-fix for the entrepreneurial city (Gaffney and Robertson, 2018;Grossi and Pianezzi, 2017;Hollands, 2015;Marvin et al., 2015;Pollio, 2016;Wigg and Wyly, 2016).Focusing on mobility, Lyons (2018: 7) similarly suggests that "smart is big business," and that it is unlikely providers will respond in a balanced fashion to the needs of the public, the public sector, and shareholders.Rather, some of these businesses have increased corporate control of urban politics (McNeill, 2016), to the point where cities run on "a top-down, master-planned vision shaped around the needs of suppliers rather than the needs of citizens" (Calzada and Cobo, 2015: 30).Similarly, companies enroll the smart discourse to strengthen their competitive position, thereby consolidating the association of smartness and digital technologies: "cities with IBM technologies and integrated systems as advised by the company are smart: those without are not" (Gaffney and Robertson, 2018: 51).Yet, evidence as to the efficacy of digital smarts is sparse while spatial and social inequities may be perpetuated and meaningful public participation in urban governance and decision-making diminished (Martin et al., 2018;Tomor et al., 2019;Wigg and Wyly, 2016).
Many of the critiques of the smart city apply to smart mobility, including smart cycling.For instance, much of the deployment of smart transportation ICTs has involved public-private partnerships that render the information collected a private commodity rather than a public good.One clear example of this relates to smart bikeshare initiatives.These not only homogenize the kind of bodies that can rent bikes within and across cities -thus generating various forms of exclusion-and lock-in only one (commercial) version of bikeshare into sociotechnical imaginaries around cycling (Nixon and Schwanen, 2019); they can also become tools for data harvesting and capital accumulation by transnational corporations (see, for example, Duarte, 2016;Spinney andLin, 2018, 2019).The privatization of these data places their interpretation, and responses to this, out of the reach of the public.
At the same time, Lyons claims that, "smart need not be technologically sophisticated … [and] being technologically sophisticated may not always be smart" (2018: 9).Tironi and Valderrama (2018) vivify this in their description of how a Chilean bottom-up cyclist self-tracking data crowdsourcing system, which was originally intended to inform government cycle infrastructure planning, lost its professed civic and democratic connotations when the engineer was forced to algorithmically "purify" the otherwise incoherent data collected.The authors conclude that the technological sophistication that comes with increasing use of digital sensors may result in the loss of not only the smartness resulting from citizen participation but also in what they call "idiotic" data, i.e., data that are incomprehensible and absurd in the context of the original project's framing or meanings, thereby indicating that such sophistication may not "fix" the issues it was believed to address.Lyons (2018: 9) contends that smart mobility should rather be "connectivity in towns and cities that is affordable, effective, attractive and sustainable," and provides a carefully thought-out example of "Walking as a Service" where digital mapping provides navigation aid alongside the affordances provided by footwear and smooth pavements, among other things (Lyons, 2020).
There are also some larger epistemological and methodological questions associated with the capture of mobility "big data."For instance, most (but not all) big data collected through digital technologies are quantitative; we may ask what types of questions these data are able to answer, or even whether the data are shaping the questions themselves.Assuming a perfect world of data collection and a positivist ontology, quantitative data can be useful in assessing what exists; but what role does it play in asking why?For instance, a "smart" approach may detect a hypothetical woman (in the aggregate) not cycling, but does it explain why she does not ride?This question is particularly relevant in a context where more advanced smart mobility assemblages, such as bike hire systems, have demonstrated inequalities along socio-demographic lines (see, for example, Corcoran and Li, 2014;Prieto et al., 2017), as in the case of the London system, where women constitute only 25 percent of the membership, 16-20 percent of trips, and 29 percent of cycle hire travel time (respectively, Beecham and Wood, 2014;Goodman and Cheshire, 2014;Woodcock et al., 2014).
Whereas the critiques of smart cities are voluminous, and those of smart mobilities and smart cycling are growing, there are currently none that intentionally re-conceptualize "smart" through a focus on the knowledge and responsiveness of human communities.

Smarts in Grassroots Initiatives that Enhance Mobility Capabilities
Academic thinking on innovation processes has traditionally foregrounded technological change, while technological innovation continues to be centered in governments' innovation.Partly in response to the developments, grassroots innovation movements have "arise [n] in reaction to perceived social injustices and environmental problems often arising in conventional innovation models" (Smith et al., 2014: 114-115).Grassroots innovations are extremely diverse, ranging from urban gardening and bicycle mechanics projects that seek to keep youth from a career in crime, to community energy initiatives, but tend to share several characteristics.These include: a focus on servicing social needs rather than generating profit (and, in fact, many grassroots innovations projects are run by non-profit charities); being animated by explicit social values revolving around, for instance, egalitarianism or localism; and being sustained through frugal, communal efforts that often demand volunteerism, grants, and exchanges (Seyfang and Longhurst, 2016).They provide novel solutions for sustainable development by using their communities' embodied experience, knowledge, and skills (Hossain, 2016), with communities understood not as internally harmonious, monolithic, and collaborative groups that neoliberal states can co-opt and responsibilize when governing social issues or processes of change, but as complex and differentiated social arrangements in which belonging and cooperation occur alongside contestation and dissent (Aiken et al., 2017).Grassroots innovations may attend to the social needs associated with this differentiation, such as the digital divide arising from socioeconomic disadvantage (Gaved and Mulholland, 2010).
While scholarship on grassroots innovations is a major source of inspiration for the research reported in this article, grassroots activity to support cycling and walking in cities can exceed the category of innovation; we therefore prefer the more generic label of grassroots initiatives.As far as cycling and walking is concerned, these initiatives are not only diverse, needs oriented, value driven, and frugal but also seek to improve people's "capability set" (Sen, 1999), a notion that is central to the Capabilities Approach conceived in development economics and political philosophy and increasingly used in transportation and mobility research (Robeyns, 2017;Schwanen, 2022).The capability set denotes a person's real opportunities to achieve outcomes that they value, or more broadly their freedom of choice to achieve desirable activities and states of being (Sen, 1999).A person's capability set may include the opportunity to become mobile, and through this find employment, reach an elder in need of care, or enjoy a distant park, for example.With respect to mobility, it cannot be assumed that availability of transportation resources alone, such as automobiles, bicycles, or walking shoes, will enable people to drive a car, ride a bike, or walk.Rather, resources have to be converted into mobility capabilities, or genuine opportunities to move through physical space, undertake valued activities en route and/or at destinations, and obtain particular experiences of/through mobility and at destinations."Conversion factors" are the conditions and processes that allow resources to be turned into capabilities, and include among others cycling skills, know-how about how to use cycling-related apps on a smartphone, the confidence to go out and about on one's own, and safe physical infrastructures for cycling and walking.Grassroots initiatives, from collective bike rides or bike mechanics workshops to the (unofficial) painting of bike lanes, provide and reconfigure the conversion factors that enable mobility capabilities.
It is not a priori clear if and how the use of digital technologies helps cycling and walking-oriented grassroots initiatives to reconfigure conversion factors and enhance mobility capabilities.Smartness is therefore less about digital technologies, data, and connectivities as the a priori means and solution to a series of ends and problems.It rather involves the contingent, pragmatic, and tactical use of such means if and when this allows the initiators and organizations of grassroots initiatives to enhance the mobility capabilities of the individuals and communities they are targeting.
Despite the promise of grassroots initiatives to support cycling and walking in large urban areas, very little research has been done on this and digital technologies.Lyons et al. (2012) frame grassroots information and communication technologies as a growing "niche" activity that may either reconfigure or transform intelligent transportation systems.Ross et al. (2012) studied grassroots innovations in which apps for sustainable mobility, such as walking and ride sharing, featured prominently.They found that the innovators were usually motivated by the frustrations that they, and those they knew, felt about existing transport options and their capacity to move.They also showed that innovators' remarkably detailed understanding of the problems they strove to resolve was heavily informed by embodied experience.The grassroots leaders in their study perceived the government and transport industry as conservative and unwilling to invest in sustainable transport, and, therefore, saw the grassroots innovations as replacing, rather than competing with, the more incremental innovations advanced by government and industry.

The Research Project
The research from which this article was developed aimed to understand who gets involved in grassroots initiatives to support cycling and walking, why they get involved, what activities and schemes are undertaken, where and why they succeed or fail, and whether they can contribute to a more equitable and just transportation system.It included 70 semi-structured interviews with leaders, staff, and beneficiaries of the activities of 49 organizations in London and São Paulo in 2016 and 2017.The present analysis references the interviews with nine research participants, but the points they make exemplify those expressed by many of the participants throughout the study.
London and São Paulo were chosen because they are global cities and exhibit several similarities and differences.Both possess strong economic gravity for their respective hemispheres and are in the top five financial centers for their regions (Z/Yen Group, 2023), and yet, significant economic inequalities exist within their respective metropolitan regions (Oyeyinka 2010& OECD 2017).However, whereas some have claimed that London has reached "peak car" (Keyes and Crawford-Brown, 2018;Metz, 2015), and both sides of the political spectrum there have demonstrated support for alternatives to the automobile, the motor vehicle fleet in São Paulo, and Brazil at large, continues to grow rapidly (da Silva, 2020;de Vasconcellos, 2005), encouraged by pro-car tax regimes and transportation planning.The larger research project to which this article owes its naissance sought to compare cycling-and walking-related grassroots initiatives between London and São Paulo, but this is not the intention here as the digital technologies were used (or not used) in such initiatives in similar ways in both cities.
Organizations that undertake grassroots initiatives regarding cycling and walking were sampled differently in London and São Paulo due to differences in data availability.In London, four boroughs were chosen using the governments' data on active transportation levels and the Index of Multiple Deprivation.Grassroot, NGOs and governmental organizations therein were then identified using Internet searches, local key informants and site visits (the latter included participation in grassroots events).As equivalent data for São Paulo were unavailable, Internet searches, local key informants and Como Anda's (2017) national survey were used to compile a list of organizations from which those that focused on particular disadvantaged neighborhoods and/or groups were selected.
Semi-structured interviews were audio-recorded and transcribed except where participant discomfort necessitated reliance on note taking only.In São Paulo, just over half of the interviews were conducted in Portuguese, and their transcriptions translated.Transcriptions were analyzed and coded using multiple interpretive readings and NVIVO software as an organizational tool.
A typology of all interviewed organizations across London and São Paulo can be seen in Table 1.This includes the social need serviced, transportation mode, target of intervention, and legal structure.Many combinations of the elements in these categories were found, such as charities who did group walks for women in low-income neighborhoods seen as dangerous, or cycle provision and training for refugees and asylum seekers.Table 2 shows the specific combinations of characteristics among the interviewees quoted in this article.

Smart With or Without Digital Technology
The grassroots initiatives in London and São Paul we studied were very diverse and ranged from group walking and cycling initiatives to bicycle mechanics workshops to the provision of digital information about the availability and use of particular transportation options.Most of those initiatives were financially lean and specifically targeted disadvantaged groups (see also Table 1), thus exhibiting the frugality and focus on social needs and social values associated with grassroots initiatives as described in an earlier section.Across the initiatives, digital technologies were used in multiple ways, including to make information available, to collect data, or to reach out to people who might benefit from participation in events like workshops and group walks or rides.In all cases the use of those technologies was not a given but contingent, pragmatic, and tactical.

Digital Social Media, Sometimes
Social media were the digital technology that was most compatible with, and useful to, the grassroots initiatives supporting cycling and walking that we studied.Just over half of the organizations behind the initiatives in London, and more than three-quarters of those in São Paulo possessed a web presence, and many suggested extensive use of ICTs in the form of social media, for both recruitment of beneficiaries and networking among themselves and broader audiences.Enhancement of mobility capabilities among the people targeted by the initiatives was an important indirect benefit of this use of social media.Digital technologies can connect bikes, helmets, and cycling bodies into broader networks and so alter velomobility (Behrendt, 2016;Nikolaeva et al., 2019), yet they can also extend the conduits of democratic communication and civic exchange associated with smart city developments (Wiig and Wyly, 2016) to cycling.Moreover, the capabilities opened up by virtue of reliance on social media may also reconfigure some of the (global) power relations in which cycling-related grassroots initiatives are embedded.Clarita is the leader of a grassroots organization that supports active mobility among children in São Paulo.To her, social media afforded an opportunity to flip the "North" to "South" flows of information about how to set up and structure walk-to-school projects, and create flows between "Southern" cities: In [name of favela] I chose two schools for the two areas, so it was working, this kind of survey … But the important thing is that in all the process, my big struggle was that I had a lot of information, methodologies, manuals, whatever, website on how to do a walk to school project in the first world [emphasis added], but no reference, references of here, the third world … I thought it would be so great to have a website with references of these projects in Latin America or India or Africa or whatever.Okay, so since it doesn't exist, I'm going to create one.That's why I created [the project] network … Because of my need of having references of urban mobility and city projects with children in Latin America.And that's the reason why it only has projects from Latin America.Only … I have [site visitors] from 50 countries from these five continents.I have a lot of public all around the world.So a little project that is really local, I put it, I publish it in the network, and it goes like whooshhhhh … So, that's the first aim, yes, to amplify the project, and second is just to inspire other people here.
Much like Ross et al.'s (2012) participants' belief (mentioned above) that they were replacing, rather than competing with, more hegemonic innovations, Clarita saw her website as a clear alternative to the discursive dominance of the Global North in literature on children's active mobility.
However, some organizations, particularly in London, did not rely as heavily, or at all, upon social media, citing a lack of human resources necessary to engage the public in this way.Two co-leaders of a cycle repair initiative for women and gender variant people in London, Linda and Mary, revealed their ambivalence during their interview: Linda: I think that's perhaps something that we could look into offering as like a website thing maybe.You know?Mary: Yeah, like a little, we're not really big on social media stuff.

Linda: No.
Mary: But people use it all the time.In this example the leaders of the grassroots initiatives felt it wiser to apply a larger portion of their finite human resources to the face-to-face activities necessary to get people cycling than to digital connections (or, mentioned elsewhere, the preparation of proposals for grants from external donors).Recruitment-which was not aggressive owing to the leaders' need to balance their charitable work with unrelated paid employment-found success through non-digital means: Mary: So I spoke to [other grassroots leaders], and they said, "Go to this refugee center [details omitted]."It would make perfect sense if I were to then pick up on that group and say "would you like to come and learn some skills?"Now, there was two women who came out of that from meeting them … And then I did spend a long time going to refugee centers and meeting people … You know what, I think the best way [to recruit] is a poster.Because I asked loads of people "how did you find out about us?" "I saw a poster."I did spend a long time going to refugee centers and meeting people.
This demonstrates community smarts as defined previously because the leaders know their audience.Mary's preference for physical posters and face-to-face visits over digital media for recruitment comes, in part, from her organization's intimate familiarity with their local target audience as well as a sense of how best to divide their time.A further potential advantage of non-digital communication in these contexts is that it avoids the digital divide that may undermine disadvantaged groups' ability to access and capabilities to use smart phones and Internet connections so as to enhance their ability to move (Durand et al., 2022;Zhang et al., 2020). 2  In summary, leaders of grassroots initiatives may or may not use digital communication tools to interact with their beneficiaries and extended networks, and the decision to do so is not a given, but rather rests on their detailed understanding of their respective communities.

Not Only Numbers
One of the primary tasks that smart sensors complete is to count-counting people moving through turnstiles, users of cycle tracks, cards being tapped, and miles cycled on shared bikes or through fitness apps-usually for the purposes of transportation modeling, financial reporting, performance management, etc.Most organizations interviewed were in support of basic pedestrian and cyclist counts as one of many sources of information that could further their efforts.In São Paulo at the time (2017), the local government did not comprehensively count pedestrian or cyclist movements; the counts it undertook tended to support private interests such as Itaú Unibanco's bike share system (Lemos, 2021).In this context, seven grassroots organizations attempted to collect this information themselves using a range of digital technologies (as is happening across the Americas, for instance in Mexico City and Los Angeles-e.g., Castañeda, 2022;Davidson, 2017).
Despite this appreciation for certain kinds of numbers, several interviewees revealed their skepticism that numbers and new digital technologies were sufficient.Jimmy, a co-leader of a peripheral London cycle repair workshop discussed this in light of the assessment practices of funding agencies, on which his and many other organizations of grassroots initiatives were dependent for their continuation: Alexandro, an employee of an organization that aids grassroots initiatives that support cycling and walking in São Paulo, said: I don't think technology will solve everything.I think it can be a tool, but there's a lot of hype associated with technology as a panacea … I think it's not just numbers.I think it's also quality.If you take quantitative and qualitative, I think sometimes organizations had to go with just a few people, go deep on something and bring very amazing things after that deep dive in some issue.
And perhaps most strikingly, Beatriz, a high-level government official in São Paulo involved in transportation but not positioned within the formal transportation authority claimed that "circulation and traffic is a qualitative activity, you need to think it culturally" [translated from Portuguese].This comment is particularly apt in light of the stories by several London grassroots initiative leaders about new immigrants and refugees, and especially the women among them.These leaders suggested that they had worked through face-to-face conversations with those in immigrant and refugee communities to figure out why women in those groups were not attending their activities.As they developed a deeper cultural understanding of why these women did not normally ride (e.g., religious edicts demanding gender-segregated physical activity), they were able to tailor programs that catered to the women's needs and found much greater participation.The women were then able to enjoy the qualitative difference that cycling made to their capability set.Echoing the discussion from the earlier section "Critical Thoughts on Smart Mobilities," the numbers garnered from digital collection may identify the quantity of beneficiaries of a project or a gender divide therein, but they may have more difficulty communicating how mental health patients have benefited or why some groups of women immigrants and refugees do not partake in supportive cycling initiatives.

The Limits of Digital Geo-Spatial Technology
Even when it is first assumed that a technological intervention will fix an issue, digital technologies may prove an inappropriate response to a particular problem.There are not only limitations to the kinds of problems that can be solved with digital data, but also with digital technologies more broadly.
A walking, rather than cycling, example from São Paulo powerfully demonstrates the way that the knowledge that is embodied in communities and accessed through practical immersion makes the use of digital technologies in grassroots initiatives contingent and provisional.Two university student women, Ana and Joana, learned that many of their peers were scared to walk alone around the university area at night: Ana: It's very difficult because women are seen [as] an easy target, so we are more vulnerable … The route to [the] metro to college in the morning was very calm, and we felt safe, but at night, that road was like terrible … We were very scared.And we start observing the movement of the girls in the college, and we saw that they were organizing some groups to walk together … So, we thought that we could build a net, that we'd connect all of that [in] real time, and in a very easy way … So we … used Uber as [an example].
After doing some basic market research they decided to create a smartphone app to match women looking for other women to walk together for personal security.They made a number of important connections with others, such as software developers, who could help them in their task.However, beta testing was not particularly successful, and the more they experimented and investigated, the more they realized that there were larger cultural influences that undermined the efficacy of their digital tool: Ana: We have a cultural barrier that woman can't be friends of another woman.
Joana: … And that's one of the reasons that the app is not available right now.
Joana: It's like we're enemies, we're not friends … What I wanted to do is to, you know the word in English 'sorority'?… So that was a word that we were trying to put in practice.And a sorority says that women are not enemies.They're friends, they're sisters, but we have this culture barrier … We have this problem here that a woman feels safer with a man, not with another woman because they say that with another woman, it's two easier targets … Ana: the known man … Joana: A friend.
Ana: A boyfriend.
Joana: A brother, a father.
They therefore decided to shift their focus from the digital to education to achieve change in "culture"-values, beliefs, dispositions to act in certain ways: Joana: We kept going for one year … We talked to each other, and we said, we can continue … Maybe it will be frustrating and everything, or we can change the route and make something different, not losing our purpose, and that's why we changed.We're not working with the app anymore, but we're working with women's rights and gender equality.
Ana: So with the app, we won't … solve the problem … That's why we need to talk with men and women.We need to teach men that they don't need to harass women to be men.Because like it is a cultural thought that we have here in Brazil.So, the best solution is to work with education.And education in … childhood.That's why now our focus is to talk with parents and help them to [raise] their children in an equitable way … This excerpt not only illustrates the limitations on the capacity of geospatial technologies to reduce inequalities in mobility capacities, but also shows the pragmatic use of digital technologies and the relevance of community smarts.Ana and Joana abandoned the app when they realized it could not serve the evolving aims of a grassroots initiative, and instead decided to rely on close and extensive dialogue with the community they hoped to help.The appropriate intelligence for the situation at hand was embedded in the community and its struggles, rather than inhering in the tool.

Community Versus High Level Spatial Data Acquisition and Ownership
Peripheral parts of Southern cities where informal provision of housing and transportation is particularly important are often disregarded by the digital mapping undertaken by big tech companies that underpins smart city systems.The resulting lacunas in digital spatial representation can exacerbate the digital and socioeconomic divides characterizing those cities (Gaffney and Robertson, 2018).However, our data suggested a complicated relationship of such companies with grassroots initiatives to support walking and cycling.Felipe is the founder and co-leader of a cycling-oriented grassroots initiative in peripheral São Paulo that, among other objectives, seeks to enhance people's cycling and walking capabilities by mapping and providing (digital) information about physical infrastructures for cycling and walking in favelas.He spoke of how their organization undertook the literal legwork (i.e., labor) for companies and government but through this were able to map infrastructures too informal to otherwise be recognized by these larger institutions: You often have lanes and alleyways that haven't been mapped.The people from Google called us for a meeting because we had started to map them using Street Maps … We had all started to map the alleyways in our neighborhood manually to get them into Street Map on Google, because we know that they eventually include them.Soon after Google had got in touch asking us to do it they created a platform for others to map areas that they weren't aware of.And now you can map them using Google.No one knows but we were a part of that.After that an American company called Map Lari got in touch asking us to take photos of the cycle lanes … We made so much stuff for free, we would ride around a lot, I have a collection of photos of cycle lanes … We've got stuff that even [the transportation authority] doesn't have.Like before and after photos, we did everything, we would do the work that they were supposed to [laughter], we would do their job for them.We had an enormous collection that would draw a lot of attention.But the cool thing about this movement … was that we wanted to map even the improvised pedestrian bridges.If there was a stream somewhere, with a big avenue going over it, but there was an improvised bridge, we wanted to map it as a pedestrian route.[translation from Portuguese] This story reveals the political economy behind the production of digital maps and illustrates how, in the smart city era, "[t]he citizen is a data point, both a generator of data and a responsive node in a system of feedback" (Gabrys, 2014: 38).But it also shows how the grassroots organization used the digital mapping apparatuses of big tech companies in a tactical manner to serve their local community rather than accepting a top-down imposition of data assumed to be locally smart.In essence, they allowed their community smarts to be incorporated in, and scaled through, corporate databases in order to realize two objectives.First, they managed to ground and situate pre-existing digital maps' "view from nowhere" (Haraway, 1988) in the favelas' infrastructure for cycling, walking, and other active mobilities (though that infrastructures' default invisibility in publicly available, privately owned maps is itself indicative of the socio-spatial relations within which corporate mapping takes place); and second, they helped to extend the mobility capabilities of those in their community.The implication is also that maximally effective digital data creation within smart cycling assemblages is not always reducible to automated data harvesting (Duarte, 2016;Spinney andLin, 2018b, 2019).
The above narrative is of course not free from capital accumulation by "big tech," facilitated by its co-optation of the voluntary and free labor of the grassroots organization, but that does not leave the latter totally without agency.A gentle "infrapolitics" (Scott, 1990) is practiced whereby the grassroots volunteers subtly subvert the tech companies' strategies and so digitize and map otherwise invisible routes for cyclists and even improvised bridges, which may provide vital accessibility improvements to some favela residents.There is, in fact, an adaptive, frugal pragmatism-common to grassroots initiatives-to the way Felipe and his organization synthesize their embodied knowledge of both their local active transportation infrastructure (however improvisational) and their beneficiaries with "free" digital communication tools.

Conclusion
This article has analyzed the use and disuse of digital technologies by grassroots initiatives supporting cycling and walking in London and São Paulo in order to complement prevailing top-down understandings of "smart" (in which recent digital technologies are taken to be intrinsically beneficial to sustainable transportation) with a community smarts perspective that foregrounds the collective and interactive intelligence of the people involved in the considered grassroots initiatives.
Our analysis of these initiatives has generated four empirical insights.First, digital communication technologies, while often useful, are not always the best means for enhancing mobility capabilities.Whereas a (slim) majority of the organizations interviewed used social media, sometimes to great effect (such as one case where dominant North-South information directional hierarchies were flipped), just under half of the initiatives relied successfully upon face-to-face networking to reach disadvantaged communities outside their networks or those without access to ICTs.
Second, in some circumstances qualitative evaluations may be just as, if not more important in building mobility capabilities than the quantitative measurements often collected through digital smart sensors.Most organizations stressed the importance of gathering qualitative information on who benefited from and the efficacy of their initiatives, alongside quantitative assessments.Quantitative measurements alone were seen to miss important understandings of how (e.g., how culture may condition mobility capabilities) and why (e.g., why a particular initiative was "amazingly" transformative despite a smaller number of beneficiaries).
Third, depending upon larger cultural contexts, digital technologies may even be severely limited in their capacity to improve mobility capabilities.An attempt to design and beta-test a smartphone app to help women to meet and walk at night together (so as to increase safety in risky neighborhoods) was deemed a failure owing to larger cultural issues around gender relations.Using the local knowledge they had developed through interactions with women pedestrians "on the ground," the creators decided instead that face-to-face education on gender equality for parents was a more effective way to cultivate women's mobility capabilities.
Fourth, mobility capabilities may be developed through the integration of local knowledge (e.g., the location of unmapped improvised infrastructures) with pre-existing abstract corporate maps first generated through automated data harvesting.Although this may be seen as the appropriation of the grassroots organization's labors and their concomitant "community smarts" by large corporations, the integrated top-down and bottom-up knowledge still ultimately increased the collective mobility capabilities of the residents in this low-income community.
Altogether, our research participants' narratives suggest that enhanced walking and cycling capabilities may emerge from the responsive knowledge of the grassroots leaders and the collective intelligence of the interacting community.Depending on the situation at hand, this may involve using, augmenting, constraining, or excluding digital technology.In all of the empirical examples provided here, the important interactions occurred between people, sometimes through the use of digital technologies (social media and crowdsourced mapping) and sometimes through their absence (physical posters, face-to-face conversations, and educational efforts).We argue that "smart cycling" or walking depends most upon the respective community's smarts, and that digital technologies be re-imagined as appendages or tools rather than the decision making "brains" behind smart mobility.

Contributions and Future Research
We contribute to the growing critique of hegemonic discourses on smart mobility by offering a nuanced analysis of both the use and rejection of digital technologies in grassroots initiatives to support cycling and walking in London and São Paulo.Our re-claiming of the term "smart" seeks to restore meanings closer to the original connotations, prior to its appropriation for corporate leverage and entrepreneurial urbanism, but also to build on these by stressing the qualitative, cultural, and interdependent dimensions of community smarts.
Embedded in the grassroots initiatives explored here is a fine-grained understanding of the often-complex mobility needs existing within their communities.This is an understanding that has been cultivated and accumulated by immersive and embodied interactions of on the one hand the leaders, staff.and volunteers and on the other hand the individuals and communities targeted by the grassroots initiatives.In the Giffinger and Pichler-Milanović definition, smart cities are, "built on the … activities of self-decisive, independent and aware citizens" (2007: 11).We suggest instead that, at least in the case of the grassroots initiatives we explored, this intelligence is founded upon interdependent citizens and collaborative decision making.As we have shown, approaches to transportation problems that rely too heavily upon digital technologies may miss important qualitative and cultural barriers to mobility.By asking "why" and "how" rather than just "what," the interviewed leaders of grassroots initiatives caught these oversights and attempted to overcome the barriers in creative ways.
We also contribute to the grassroots innovation literature by demonstrating how grassroots initiatives tend to cultivate capabilities in the manner that is understood in the Capabilities Approach (Robeyns, 2017).The focus here has been on capabilities for cycling and walking, which offers an environmentally and socially sustainable way of addressing transport poverty and allowing disadvantaged individuals to live the lives they value.The literature on grassroots initiatives in urban mobility is unduly limited, and this article joins a handful of others addressing this oversight.
The research project upon which this article is based was relatively exploratory and employed a fairly broad set of interview questions.Future research might focus more specifically on the sources of knowledge (or "intelligences") that undergird grassroots initiatives in urban mobility and the role (or intentional avoidance) of digital technologies in their creation and operation.

Policy Recommendations
Public funding for smart mobility objectives tends to support further developing digital technologies rather than sustaining or expanding grassroots initiatives, despite the profits that already accrue for private sector smart interventions.Between 2012 and 2021, the UK national government invested £5 billion and £93 million into, respectively, ICTs and digital technologies to support smart cities (UK POST, 2021).Owing to devolution, it is difficult to estimate how much national funding supported grassroots community initiatives during this time, but between 2015 and 2023 Transport for London (which in addition to transit fares receives some money from higher levels of government) provided roughly £2 million through its walking and cycling grant for community organizations (TfL Walking and Cycling Grants, 2023).The disparity is striking.
We recommend that grassroots organizations that substantively support walking and cycling-and particularly those targeting disadvantaged communities-receive a larger portion of the monies distributed for transportation improvements even if, under austerity measures, this must come at the cost to subsidies for mobility ICT development.We also recommend that use of the term "smart" be applied, when appropriate, more broadly to mobility interventions that do not necessarily rely upon digital means.
Whereas digital interventions have their place in building better (i.e., more sustainable and just) transportation systems, their proponents should not have a monopoly on the term smart, nor a monopoly on funding.

Table 1 :
Typology of organizations studied

Table 2 :
Characteristics of the organizations quoted in this article Speaking for myself, I recognize and I appreciate and I acknowledge the value of measuring what has been done.So I'll go back to my previous statement.What's your criteria?What do you use to evaluate?… How do you evaluate what's going on?Some things you can measure straight away.If somebody said to us, "take 20 people, take them cycling, and see if they lose weight;" that's a measurable outcome.But then if you've got 20 mental health patients, and you take them cycling, how do you measure their mood?How do you measure how they feel?How do you numerically quantify it?Do you numerically quantify?