Grassroots innovations are a relatively new cousin of the Strategic Niche Management (SNM) and Multi-Level Perspective (MLP) analytical frameworks for theorising low-carbon transitions. They are most often defined as, “a network of activists and organisations generating novel, bottom-up solutions for sustainable development and sustainable consumption” (Hossain, 2016; Seyfang and Smith, 2007: p. 585). GIs are different from conventional low-carbon innovations in several distinct ways.

First, they aim to service local social needs (Blake and Garzon, 2012; Hossain, 2016; Reinsberger et al., 2015) and support broad social change (Hess, 2007), rather than generate profit. In this way GIs are strongly value driven (Bradbury and Middlemiss, 2015; Gupta et al., 2003; Haxeltine and Seyfang, 2009; Hielscher, Seyfang and Smith, 2011; Martin, Upham and Budd, 2015; Pansera and Sarkar, 2016; Ross, Mitchell and May, 2012; Seyfang, 2009; Seyfang and Haxeltine, 2012; Seyfang and Smith, 2007; Smith, 2006; Smith et al., 2016), with social, cultural, and ethical values different from mainstream sources of innovation (Monaghan, 2009). These broader social values not only drive the innovations but are themselves said to be extemporaneously and iteratively negotiated through the process of innovation (de Vries, Boon and Peine, 2016; Grabs et al., 2016; Nixon and Schwanen, 2018; Wolfram, 2018). For example, new residents from an ethnically unfamiliar background may prompt the expansion of the geographic and demographic extent of an organisation’s domain of concern (Nixon and Schwanen, 2018). The values held by grassroots leaders and beneficiaries may be diverse, even tacitly conflicting, yet still overlap well enough for productive collaborations, such as with co-existing members of free re-use groups who may range from radical anti-capitalists to frugal conservatives (Martin and Upham, 2016; Seyfang and Longhurst, 2013; White and Sterling, 2013; Wolfram, 2018). The process of negotiation towards mutual understandings itself may beneficially challenge and alter normative values (Grabs et al., 2016).

Second, and borne of these values, those involved in GIs often take an inclusive stance in both the process of creating, and subsequent use of, an innovation. They thus often focus on marginalised groups. Some initiatives also include people from these groups among the participants and/or leadership (Amin, 2005; Geels, 2012; Gupta, 2012; Kirwan et al., 2013; Pansera and Sarkar, 2016; Seyfang and Smith, 2007; White and Stirling, 2013).

Third, they involve locally responsive experimentation to produce novel bottom-up solutions to environmental and social problems, and communal efforts that often demand volunteerism, grants, and exchanges (Hossain, 2016; Nixon and Schwanen, 2018; Seyfang and Longhurst, 2016; Seyfang and Smith, 2007). Strong understanding of need is not akin to knowledge of how to proceed, so these initiatives are often “living experiments” that do much of the trial and error work too onerous or unprofitable for larger organisations. They could thus be seen as “path finders” of socially just and sustainable transport (Aiken, 2015), developing promising praxiologies (Kotarbinski, 1965) for transitions.

Grassroots innovations may sometimes overlay but should not be considered synonymous with mainstream activism. As suggested above, participation in GIs is not exclusively dependent upon pro-social values (Martin and Upham, 2016), though they are often involved, and some innovators intentionally avoid the overt politics of mainstream activism. But as Chen, Lune and Queen (2013) argue, organisations reflect and propagate values, and GI initiators could be said to “lead by example” in enacting their values through concrete projects. Chatterton and Pickerill (2010) describe a “prefigurative politics” as one in which the means actively shape the ends, rather than traditional political consequentialism where the ends justify the means. This may also reveal existing deficiencies in a mainstream system, such as automobility. As Smith et al. (2014: p. 122) argue, “by trying to do things very differently, and in coming from or engaging with the excluded, GI movements make very visible the institutional, political and economic injustices of conventional innovation systems” (see also Smith and Seyfang, 2013). In this way grassroot innovations could be said to intentionally or unintentionally exemplify, or provide a “prototype” for, alternative practices and politics that support diversity and inclusion.

There is an arguable paucity of studies of mobilities-related grassroots innovations (Ross, Mitchell and May, 2012). Ross, Mitchell and May (2012) analyse the struggles and successes of sixteen grassroots transport ICT innovators—largely mobile phone application designers—using user innovation and social innovation theories. Schwanen (2015) describes the way the different innovation contexts of Oxford and Brighton uniquely influence the pathways of the cities’ respective GIs in low-carbon mobility. Although not framed as grassroots innovations, de Boer and Caprotti (2017) describe the importance of GIs to the cycling transition in The Netherlands, such as the white bicycle free sharing scheme, and the Stop de Kindermoord organisation’s reclaiming of automobilized space through such activities as do-it-together bicycle lane paint. Though framed more as grassroots campaigning, the activists described by Newman et al. (2008) deployed temporary street closures, at first informally, in an attempt to create pedestrian friendly spaces in a Toronto neighbourhood.

Supportive infrastructure is a requirement for the equitable expansion of active transportation, but what exactly constitutes “infrastructure” is currently under theoretical interrogation. The term itself is derived from the Latin infra, which means “below” and the more obvious “structure” (from French); in other words, the structures that support a system. To think of only physical infrastructure and its enchanting promises of progress and modernity (Harvey and Knox, 2012) would be to miss other critical supportive systems and affective dimensions of infrastructure (Cupples and Ridley, 2008; Larkin, 2013). Liang et al. (2014: p. 1135) found that “…an objectively good environment for bicycling was necessary but not sufficient for [encouraging] bicycling”, and as Star (1999), Larkin (2013), Amin (2014), Schwanen and Nixon (2019), and others have claimed, we need to think more broadly about what constitutes infrastructure, particularly its social and psychological dimensions. In other words, there is a need to broaden the narrow focus on the physical and material elements of walking and cycling infrastructure (Lugo, 2013; Temenos et al., 2017).

Lugo (2013: p. 206, emphasis added) contends that, “instead of reducing movement in the street to an individual engagement with physical transport infrastructure, the concept of human infrastructure emphasises the role of social interaction in how people move… People are part of the infrastructure enabling or disabling certain mobilities”, such as when high cycling rates reduce collision risk and thereby facilitate greater cycling mode shares (Jacobsen, 2003). Bicycle mass rides like CicLAvia in Los Angeles (Lugo, 2013), or street closures for pedestrians, cyclists, and leisure activities on Avenida Paulista and the Minhocão in São Paulo (Nixon, 2025), are examples of bodies refashioning the meanings and purposes behind existing physical infrastructure through collaborative efforts. Lugo argues that alternative social perspectives and associated practices already co-exist with absent or problematic physical infrastructure, creating new infrastructural realities, such as networks of cyclists sharing their embodied knowledge of best routes, advice that may contradict more hegemonic resources such as generic online maps aimed at drivers (see also Nixon and Schwanen, 2024).

The concept of “social infrastructure” has varied definitions but here it follows the first two of four concepts identified by Latham and Layton (2022: p. 660), namely: “people as infrastructure”, and particularly, “the sociality that gathers around conventional hard physical infrastructures… [and] allows otherwise precarious technological networks to function… [via] intricate social workarounds… in underfunded and under-supported urban environments.” This paper thus uses “social infrastructure” to mean networks of collective practice that both support social life and “actuate” physical infrastructure otherwise dysfunctional for certain groups of people.

It is essential to ask for what and for whom a walking or cycling infrastructure is provided, who benefits from a particular intervention, and what assumptions about needs and embodiment may be built into the design.

A good example of exclusion through well-intended yet arguably fetishized infrastructure provision is shared-space design. Shared-space is an approach that assumes that mixing pedestrians, cyclists, and drivers in the same space, with only subtle cues to subconsciously encourage polite negotiations of movement, will encourage higher pedestrian usage and more interactions that support social cohesion and community building. However, Imrie (2012) argues that these spaces are predicated upon an assumed occupation by only stereotypically-abled bodies in possession of all sensory faculties, particularly the perfect vision required to negotiate movement with others behind windshields who may themselves not see a pedestrian. In this way, the infrastructure standardisation (Schwanen and Nixon, 2019) shared spaces impose belies the “open” connotations associated with their design. Their construction to some extent represents a post-political, technocratic attempt to solve urban problems through design borne of values functionally uncontestable in their ambiguity and ossification in policy (Imrie, 2013). A result is that the needs of vision-impaired individuals are ignored, and thus they are dis-abled by shared-space design (Imrie, 2013). Yet, a car and wheelchair trip to a shared-space destination may be easier to traverse owing to fewer physical obstacles. The key lessons here are that (a) the diversity of disability is myriad, and involves both physical and psychological dimensions; and (b) capabilities, including motility, are determined at the “interface” between the individual and their socio-physical environments, not strictly in one or the other.

Historically, disabled people have been overlooked in UK cycle planning (Andrews et al., 2018; Hickman, 2016) and almost totally ignored in São Paulo. There exists a widespread assumption that disabled people do not cycle, but rather predominantly use public transit, walk, or drive, even though in many instances cycling may actually be easier than walking or taking transit (Andrews et al., 2018; Inckle, 2020). In the rarer cases where it is thought that they do cycle, it may be assumed that specially adapted cycles are always a necessity, which may not be the case (Andrews et al., 2018; Inckle, 2020). It may also be assumed that adapted cycles are all that are necessary, when this may overlook other needs such as suitable cycle parking, route directness, co-pilots, and/or information such as maps (Andrews et al., 2018; Inckle, 2020).

The social model of disability (Oliver, 1983) sees social exclusion as arising from a “failure of society to plan inclusively for a range of individual characteristics” (Andrews et al., 2018: p. 148). Dokumaci (2020: p. S99) provides the example, among many, of the twist-off cap as, “a materialized form of ‘ignorance’ of corporeal variations and vulnerabilities”, but suggests that “people as affordances” may in some cases bridge the gap. Affordances are opportunities for action arising from the relationship between an individual (or other organism) and their environment (Gibson, 1979). In this way, potential action is not a product of a person’s characteristics or an environment’s properties alone. Dokumaci (2020: p. S99) argues that disability, “can be defined ecologically as a shrinking of the environment and its available affordances… but this shrinkage can also incite invention and improvisation that makes the same environment afford otherwise”, whereby, “people may enable, facilitate, or interfere with and disable the emergence of affordances for one another” (Dokumaci, 2020: p. S98) through what Bloomfield, Latham and Vurdubakis describe as “collective accomplishments” (2010: p. 419). People as affordances thus arise, “through the complementarity of multiple perceivers and the world”, in what Dokumaci calls “care intimacy” (2020: pp. S102, S103) whereby one person intimately understands another’s access needs. Where these people as affordances come together in supportive constellations, they may be considered social infrastructure.

The literature discussing grassroots innovation and the critical scholarship on infrastructure have not, as of yet, explicitly collided. Grassroots innovations in infrastructure may be seen as “autoconstructions” (Amin, 2014: p. 155), “pioneering citizenship” (Holston, 2008: p. 8), or “people as infrastructure” (Simone, 2004: p. 407), characterisations that strongly resonate with Dokumaci’s (2020) people-as-affordances. Tonkiss (2015: p. 389) suggests that “human bodies are the basic carrying element of these auto-infrastructures… [that] fill in for, compete with, or extend incomplete, exclusionary or inadequate networks of hard infrastructure which are often more vulnerable to technical failure, environmental risk, elite colonisation or economic collapse”. Some similarly argue that in economically challenged regions, “access to services is not determined via connections to formal networked infrastructures… [but] is the emergent outcome of myriad socio-technical assemblages created by households and communities” (Goodfellow, 2016: p. 27). By combining the frameworks provided by the GI and affordances literatures, and the more critical and political interrogations of infrastructure, we can better articulate how civil society may potently contribute towards more equitable and sustainable mobility futures.

Almost all of the organisations studied catered to particular social groups with a mobility need, or more than one need, that remained unattended by the existing transport system (please see Figure 1 for a typology of the organisations studied). These groups included: disabled people; women and the gender variant; refugees and asylum seekers; ethnic, cultural, and/or religious minorities; those in deprived neighbourhoods, particularly those with lower levels of income or education; and disadvantaged age groups such as the young or old. A couple of groups served all those interested in walking and cycling in greater London and city of São Paulo, with no particular target. However, most of these groups found that the needs they served went beyond the mobility itself, and included other related necessities, such as socially supportive environments that enhance wellbeing (Schwanen and Nixon, 2020).

All of the organisations provided supportive infrastructure for walking and/or cycling in the form of: mobility training, such as learning to cycle; group mobility, such as gathering groups of potential GI beneficiaries to partake in walking or cycling collectively; cycle repair training, such as basic maintenance classes or full mechanic’s training; and signage and navigation, such as posting home-made signs, painting (guerilla) crosswalks/zebra-crossings, or designing and attempting to disseminate creative route maps. However, again, the support provided sometimes exceeded this, including such things as gifting helmet and lock kits for refugees, or tea and biscuits for the disabled and women and gender-variant groups. The organisers claimed that these “extras” rendered their activities more attractive to beneficiaries and filled in gaps in their motility and social lives.

Most of the organisations identified were small, with fewer than five full-time volunteers or employees, and although GIs may be performed by private business or social enterprise (Seyfang and Smith, 2007), most of the research participants were involved in charitable or cooperative organisations, some with social enterprise branches or goals to have revenue-generating branches. For some publications on other topics from this project, please see: Nixon (2025), Nixon and Schwanen (2018), Nixon and Schwanen (2019), Nixon and Schwanen (2024), Schwanen and Nixon (2019), and Schwanen and Nixon (2020).

The most obvious physical infrastructure that the London organisations provide are fleets of specialised cycles to be ridden during group cycling sessions. Many of the beneficiaries require cycle technologies that are different from the common standard two-wheeled “safety bicycle”, such as three or four-wheeled bikes, hand-pedaled drives, or special seats. For those that may still need accompaniment, in-line or “side-by-side” tandems, and cargo trikes with platforms for wheelchairs provide a means to cycle (see Andrews et al., 2018, Norcliffe et al., 2022; Clayton, Parkin and Billington, 2017 for some typologies of these). These cycles may be very expensive—as much as £7500 to purchase (more for electric assist) or £15–£20 per hour to rent—and thus otherwise unaffordable for many disabled people (Clayton, Parkin and Billington, 2017). Emily thus charged £5 for a session when the beneficiary’s sponsoring organisation—such as a school or religious institution—can afford it. They are also somewhat cumbersome to store; one organisation filled thirteen shipping containers at their three locations with their cycles.

The London organisations provided most of their cycling opportunities at sports stadiums and similar locations where the riding could be performed in an off-road, traffic free environment. These environments provide safe environments for the participants, and as suggested by one reviewer, also provide a respite for the oft-stressed caregivers.

In addition to finding donations of improved wheelchairs and special cycles for those with inadequate “wheels”, Lucas has established a growing spatial database of São Paulo route and destination accessibility information and two public facing “apps” to access this information. Destinations include not only stationary entities such as restaurants and museums, but also events such as plays and musical performances. Although the devices, and electrons that constitute the database and apps could be described as physical, the accessibility information inputted is gathered through social outings by a team of differently abled volunteers Lucas calls “accessibility ambassadors [embaixadores da acessibilidade]”.

Therefore, it could be said that these different physical infrastructures are actuated by social infrastructure such as: the space-time coordination of the spaces, such as sports tracks for the beneficiary sessions or the ambassador outings and new wheelchair drop-offs; the paid human mechanical intervention to adjust the cycles for different bodies and occasionally do repairs or outsource welding; riding skill training by staff; treasury accounting, admission and refreshment table volunteering; the management of volunteers and staff; and all of the “extras” that enrich the benefits for those who partake. Emily describes some of the extras her organisation provides:

These social “extras” may actually be necessary incentives to become mobile, as Lucas argues when describing the group mobility activities his organisation leads:

The challenge of getting some disabled people to leave their homes may be exacerbated by lower employment rates and larger portions of those over 60 among those with disabilities (Andrews et al., 2018).

Elizabeth alluded to the socially and physically aggregate nature of the diverse infrastructure they provide:

Without these tailored cycle “products”, and the “service” (Sagaris, 2015: p. 70) offered by these grassroots organisations, many would go on to assume that those with disabilities simply cannot cycle.

The organisations’ leaders, beneficiaries and websites reported on a host of positive impacts their infrastructure had brought about.

Besides positive direct impacts, these grassroots innovations contribute to larger transitions by setting infrastructural precedents which may be drawn upon for other interventions. They generate this new infrastructure through a process of experimentation. It is common for GI leaders to have to carry the burden of risk (e.g. financial) of experimentation in the early stages of GI growth (Ross, Mitchell and May, 2012). High risk experimentation is ongoing in the GIs explored here, as expensive cycles and shipping containers, pricey app development, or labour-intensive database building must be attained and deployed to assess their real usefulness to their beneficiaries:

Aiken argues that, “it is this experimenting with existing technology, in tweaked social settings, that alters the relationship between the infrastructure, practices, and habits” (2015: p. 768). Experimentation leads to learning and demonstration effects that increase the momentum of alternative approaches to mobility (Temenos et al., 2017). In this way the GIs “lead by example”, generating procedural and technical knowledge along the way, and thus differ in political praxeology somewhat from mainstream activism.

One of the leaders avoids formal politics for fear that this may distract her from her infrastructure provision:

As Chatterton and Pickerill argue, “resistance is perhaps better seen as hard-wired into the productive daily acts of project building rather than as a part of direct action and confrontation” (2010: p. 482).

However, Chatterton and Pickerill (2010) also claim that political identities emerge through mundane practice rather than pre-exist, and many of the grassroots leaders interviewed in the larger project spoke of how their early experimentation solving walking or cycling barriers grew into larger political interests and engagement. In this way “infrastructural subjects” (Tonkiss, 2015: p. 384) become politicised “infrastructural actors”. Elizabeth and Lucas have led their teams into far more explicit political engagement:

Lucas also explicated the politics inherent in the way his app revealed deficiencies in the government’s provision of accessible infrastructure.

The embodied experience of the aforementioned infrastructure also contributes to their political potency. All three leaders felt that the immersive experience of mobility was essential in convincing both disabled folks, and the public at large, that cycling is a realistic option for many of those otherwise considered immobile or only “automobilizable”. Elizabeth states that:

More broadly, these group mobility sessions could be said to set a precedent for the capacity of people with disabilities to cycle, given the right infrastructure. Politically, this turns the attention towards the omission of these types of initiatives in other locations, and concomitant exclusionary standardisation of walking and cycling infrastructure at large.

These organisations have thus formed two political orientations—one that quietly enacts the future they wish to see, in the form of special cycle provision and training, and the other that engages the existing transport regime directly through campaigning. But they maintain that their work is still very much about concrete, mundane mobility. Whereas Chatterton and Pickerill (2010) call this “prefigurative politics” it is suggested here that it may also be framed as the demonstrative politics of precedential infrastructure.