The Digital Infrastructure of Collective Efficacy.

Operational fragmentation in frontline security: the case for a shared communication layer.

This paper examines whether the operational isolation of frontline security workers, when considered alongside contemporary frameworks of collective efficacy and growing evidence on lone-worker mental health, constitutes a structural gap in the United Kingdom’s public safety infrastructure that warrants policy and product attention.

The Security Industry Authority records in excess of 400,000 active frontline licences across the United Kingdom^[1], with comparable workforces deployed across the European Union, the United States, the Gulf, and the Asia-Pacific region. Studies of this workforce have consistently noted three interrelated characteristics: a high rate of repeated exposure to violence and verbal abuse, a working environment dominated by lone-working assignments, and an absence of cross-employer communication infrastructure^[2].

Talas, Button, Doyle and Das (2021) reported that approximately 40 per cent of UK security operatives surveyed presented symptoms consistent with post-traumatic stress disorder, with elevated rates of persistent traumatic stress associated with sustained occupational exposure^[3]. Subsequent industry survey evidence has corroborated these findings, with respondents reporting elevated incidences of physical assault and verbal abuse within the preceding twelve months^[4].

The literature on collective efficacy, originating with the work of Sampson, Raudenbush and Earls (1997), suggests that informal social control and willingness to intervene on behalf of the common good are associated with measurably lower rates of community violence^[5]. The argument advanced in this paper is that a structurally analogous mechanism may be available to the dispersed frontline workforce, provided that the underlying digital infrastructure permits real-time, peer-to-peer awareness across employer and organisational boundaries.

The paper proceeds in fifteen sections. It situates the proposition within the anthropological and sociological literature on collective vigilance; reviews the empirical evidence concerning lone-worker welfare; examines the limitations of existing legacy systems including ACT, the Community Security Trust, Shopwatch radio networks, Project Griffin, and contemporary enterprise platforms such as Zinc CityINTEL; and considers the regulatory environment shaped by the Terrorism (Protection of Premises) Act 2025 and the Health and Safety at Work Act 1974. It then sets out an architecture, the adjacent “watchers” network thesis, six representative use cases, the international comparative landscape, a UK launch framework, and a structured mapping of investor archetypes.

400,000+Active SIA frontline licences in the United KingdomSecurity Industry Authority, 2026

40%Operatives presenting symptoms consistent with PTSDTalas et al., Policing & Society, 2021

21M+Estimated global private security workforceILO and industry estimates, 2024

Collective vigilance as a recurrent feature of human social organisation.

Before considering the contemporary case, it is useful to situate the proposition within the longer historical and theoretical record. Distributed observation and informal mutual aid are not modern inventions; they are recurrent features of how human communities have organised safety in the absence — or alongside — the formal apparatus of policing.

Anthropological accounts of small-scale societies have long noted shared sentinel behaviours, including communal night-watches and reciprocal warning systems. Dunbar (1998) advanced the social brain hypothesis, suggesting that the unusual size of the human neocortex may have co-evolved with the cognitive demands of cooperative defence and dense social coordination^[6]. Whether or not one accepts the strongest form of that argument, the empirical record across hunter-gatherer and early agrarian societies indicates that collective observation has been a persistent organisational feature.

The medieval English institution of the “hue and cry”, codified by the Statute of Winchester in 1285, formalised a community-wide obligation to pursue a suspected offender on the call of any witness. Comparable arrangements appear across jurisdictions, including the Anglo-Saxon tithing system and various continental analogues. The town crier and the parish constable represented successive professionalisations of what had previously been a collective civic responsibility.

With the establishment of the Metropolitan Police in 1829, the United Kingdom moved towards a model in which the formal apparatus of the state assumed primary responsibility for public safety. As Loader and Walker (2007) observe, this transition did not eliminate the underlying need for distributed civic awareness so much as it reorganised the formal channels through which it operated^[7].

Jacobs’ observation, although developed in an urbanist context, aligns with a body of subsequent empirical work in which the density and quality of informal observation has been associated with variations in community safety outcomes. The conceptual through-line from these earlier observations to the contemporary literature on collective efficacy is reasonably direct.

In their 1997 study published in Science, Sampson, Raudenbush and Earls examined data from 343 Chicago neighbourhoods and reported that variation in collective efficacy — operationalised as a combination of informal social control and social cohesion — accounted for a substantial share of the variation in violent crime rates, after controlling for structural disadvantage^[5]. Subsequent replications across multiple jurisdictions have generally supported the central finding, although with methodological qualifications regarding measurement and causal inference.

The salient point for the present discussion is that collective efficacy, as conceived in the literature, is not a property of individual actors but of the network of relations between them. Translating this into operational terms, the question becomes whether, and through what mechanisms, a dispersed workforce of frontline observers can be enabled to function as a coherent network rather than as a series of operationally isolated individuals.

The occupational mental-health profile of frontline security operatives.

A growing body of empirical work suggests that the psychological burden borne by frontline private security operatives is comparable in several respects to that of state policing personnel, although the institutional welfare apparatus available to the two workforces differs markedly.

Talas, Button, Doyle and Das (2021), in a peer-reviewed study published in Policing and Society, surveyed 754 UK security operatives and reported that approximately 40 per cent had been physically assaulted within the preceding twelve-month period, with a comparable proportion presenting symptoms consistent with post-traumatic stress disorder^[3]. Subsequent industry surveys conducted across comparable practitioner samples have reported broadly converging prevalences of physical and verbal aggression over the same reference period^[4].

Cogan (2024), writing in Healthcare, has argued for the conceptual utility of distinguishing “persistent traumatic stress” from the single-event aetiology that frames much of the PTSD literature^[8]. The argument, broadly summarised, is that repeated low-intensity exposure to hostility — characteristic of door-supervision, retail security, and similar deployments — may generate a clinically distinct symptom profile that is not fully captured by existing diagnostic frameworks.

The lone-worker context plausibly mediates these effects. In the absence of co-located colleagues with whom to debrief, the routine decompression mechanisms that mitigate occupational stress in team-based settings — a brief verbal exchange, a confirming nod, shared situational sense-making — are largely unavailable. Whether digital channels can substitute, in part, for that absent co-presence is an empirical question that the literature has not yet resolved.

A systematic review by Shalaby and Agyapong (2020), published in JMIR Mental Health, examined evidence on peer support across mental health contexts and reported associations with improved hope, perceived recovery and empowerment, alongside reductions in service utilisation^[9]. The transferability of these findings to occupational settings is not direct, and further work is required before strong claims can be made about workplace peer-support interventions delivered through digital channels.

With that caveat noted, the available literature is at minimum consistent with the proposition that a low-friction mechanism for peer awareness and acknowledgement may have welfare value independent of any direct safety effect.

Existing reporting frameworks: an architectural review.

The United Kingdom currently operates a substantial number of overlapping reporting and information-sharing initiatives, each addressing a particular subset of the safety problem. A review of the available material indicates that, despite their individual merits, these systems share architectural features that limit their capacity to support real-time peer awareness across employer and organisational boundaries.

The UK government’s Action Counters Terrorism (ACT) campaign instructs the public and professionals to report suspicious activity via a confidential phone line or an online web form. The Community Security Trust (CST) operates a comparable email and telephone reporting model for incidents affecting the Jewish community. The British Transport Police “See It. Say It. Sorted.” campaign uses SMS to 61016 for transport-related concerns. The architecture of these systems is broadly similar: one-way submissions to a central authority, with limited feedback to the reporter and no native mechanism for alerting other proximate observers.

Two-way radio schemes (Shopwatch and adjacent networks)

Two-way radio schemes — most commonly trading under the Shopwatch, Pubwatch and TownLink banners — remain the dominant cross-employer communication channel in many UK Business Improvement Districts. The model has acknowledged virtues, including reliability and an established compliance envelope under Ofcom licensing. Its principal architectural limitations, for the purposes of the present argument, are that membership is gated by handset distribution and that interconnection across districts is not native to the design.

Enterprise platforms (Zinc CityINTEL and analogues)

Enterprise communication platforms, of which Zinc Systems’ CityINTEL deployment with the City Security Council and the City of London Police is a representative example, provide sophisticated tools for corporate security teams. The procurement model is employer-led: organisations purchase licences and provision them to named operatives within their own command structure. The operational gap that remains is the connection between operatives whose employers have not contracted with the same platform, which in practice describes the substantial majority of the workforce.

Project Griffin and its international analogues

Project Griffin was launched in April 2004 by the City of London Police and the Metropolitan Police as a counter-terrorism awareness initiative for the private security sector. Contemporary accounts of the response to the 7 July 2005 London bombings credit Griffin-trained personnel with materially supporting the emergency services in the immediate aftermath. The model has since been adapted internationally as Project Guardian (Singapore), Project Servator (UK rail and aviation), NYPD Shield, and various Australian and Canadian equivalents.

The training-led design has clear merits and remains a valuable component of the wider system. The observation relevant to the present argument is that training of itself does not create a persistent communication channel between trained operatives once the training day concludes.

An alert-first communication layer for the dispersed frontline.

The architecture proposed in this paper is not intended to displace existing reporting frameworks or enterprise platforms. It is intended to occupy the operational gap that those systems do not currently address: the connection between proximate frontline operatives across employer and organisational boundaries.

Sustained adoption of new workplace tools is well documented as dependent on integration with existing routines rather than the introduction of additional ones^[10]. The frontline security shift is structured around a small number of recurring behaviours: shift sign-on, periodic lone-worker welfare check-ins, patrol-checkpoint logging, incident reporting, and handover briefing. The proposition advanced here is that an alert-first communication layer can plausibly serve as the canonical surface for several of these activities.

Where deployment circumstances trigger duties under the Health and Safety at Work etc. Act 1974 and the Management of Health and Safety at Work Regulations 1999, periodic welfare confirmation is operationally non-discretionary^[11]. A communication tool that incorporates compliant welfare-check functionality occupies a position that is both habitual and, in effect, professionally required.

The platform does not, and is not designed to, provide a response service. The function of the network is the generation and distribution of awareness. Operational response remains the province of the existing actors — employers, dedicated control rooms, the police service and the wider emergency response system.

Adjacent occupational groups: the case for a wider network.

The licensed private security workforce is the most concentrated, regulated and operationally relevant population for an initial deployment of the kind described in this paper. It is, however, neither the only population whose work places its members in repeated and structured contact with the public, nor the only population for which a peer-awareness layer would plausibly carry welfare and operational value.

A number of platform operators — including Uber, DoorDash, Bolt and Careem — have implemented in-app safety features, in several cases through partnerships with RapidSOS for direct routing to emergency services^[12]. These features are appropriately scoped to the contracted relationship between the platform and the worker; they are not designed, and do not function, as a peer-awareness layer between workers operating for different platforms or with adjacent frontline workforces.

On reasonable estimates, the United Kingdom alone supports of the order of 1.4 million workers across licensed security, taxi and private hire, and on-demand courier roles, before any inclusion of retail, hospitality and transport categories. The addressable population for a peer-awareness network of the kind described plausibly extends, at the upper bound, towards seven million individuals.

Indicative deployment scenarios.

The architecture proposed in this paper supports a relatively wide range of deployment scenarios. The cases set out below are illustrative rather than exhaustive, and are intended to indicate the geometric and operational diversity that a geofenced peer-awareness layer can plausibly accommodate.

What we propose to measure, and what a connected frontline workforce could do.

A research argument for an unbuilt system is only as strong as the impact it is willing to be tested on. This section sets out the welfare, operational and public-safety effects we contend an alert-first peer network could produce, the measurement framework we would adopt, and the broader analogy to other crowdsourced urban-intelligence systems.

Mobilisation in crisis: evidence from analogous systems

The proposition that a distributed network of frontline workers can be mobilised at speed in a serious incident is not speculative. It has been observed in several relevantly similar systems already in operation.

The cautious reading of these cases is that a free, peer-to-peer communications layer adopted by a working population, when combined with a credible institutional partner on the receiving side, can produce situational awareness at a scale that the institutional partner cannot generate on its own.

Regulatory environment and design constraints.

The architecture proposed in this paper sits at the intersection of three distinct regulatory regimes: counter-terrorism preparedness, occupational health and safety, and personal data protection. Each imposes constraints that shape, rather than incidentally affect, the design of the platform.

The Terrorism (Protection of Premises) Act 2025

The Terrorism (Protection of Premises) Act 2025, commonly referred to as Martyn’s Law, received Royal Assent in 2025 with full commencement scheduled for April 2027^[1]. The Act places duties on responsible persons for qualifying premises and events. Standard tier premises (expected attendance 200–799) must put in place public protection procedures. Enhanced tier premises (expected attendance 800+) must additionally implement and document public protection measures.

Occupational health and safety obligations

The Health and Safety at Work etc. Act 1974, together with the Management of Health and Safety at Work Regulations 1999, imposes on employers a duty to ensure, so far as is reasonably practicable, the welfare of employees. The Health and Safety Executive’s guidance on lone working identifies private security operatives as a category warranting specific consideration^[11].

Data protection and information governance

The platform is designed in accordance with the principles set out in the UK General Data Protection Regulation, with particular attention to data minimisation, purpose limitation and storage limitation. Location data is captured at the time of an alert and at the spatial resolution necessary to define the relevant geofence; persistent background tracking is not implemented. Photographic content and free-text descriptions associated with alerts are subject to automatic time-bound retention. Subject rights, including the right of erasure and the right of access, are operationalised through standard subject request workflows.

The Citizen application in the United States provides a relevant reference point. It has demonstrated material capacity to mobilise public attention during crises, and equally has attracted documented criticism for the amplification of unverified content^[13]. The design choices proposed here differ in several respects that we suggest are material to managing those risks: posting authority is restricted to verified frontline professionals; severity classification is human-initiated and machine-validated; and there is no public consumer feed.

International outlook and market sequencing.

The operational problem to which the proposed architecture responds is structurally similar across most advanced economies, although the regulatory and commercial pathways through which a deployment would proceed differ materially by jurisdiction. The discussion below sets out an indicative sequencing of priority markets, with the regulatory or commercial wedge that we judge most likely to support initial traction in each.

United Kingdom: institutional landscape and partnership map.

The institutional environment in the United Kingdom is, on a number of dimensions, well suited to an initial deployment of the architecture proposed in this paper. The discussion that follows sets out the principal institutional partners whose remits are most directly relevant, without making representations as to current or prospective partnership arrangements.

Capital landscape and aligned investor mandates.

The proposition advanced in this paper sits at an unusual intersection of investment categories — public safety, dual-use infrastructure, frontline-worker welfare, and consumer-grade network products. The discussion that follows surveys the institutional capital sources whose published mandates and recent activity most closely align with the proposition, without making representations as to current or prospective investment interest.

The investors listed below are drawn from the public record of their stated mandates and recent portfolio activity at the seed and Series A stages. Inclusion is intended to indicate likely thematic alignment rather than to imply any particular relationship.

United Kingdom & Europe

United States

Gulf & Sovereign

A measured proposition, openly stated.

The argument advanced in this paper is, in summary, that the United Kingdom’s frontline security workforce currently operates without a persistent peer-awareness layer that crosses employer boundaries; that the absence of such a layer carries measurable welfare and operational costs; and that an alert-first, geographically scoped communication architecture offers a plausible, if not yet empirically settled, response.

We do not contend that the proposed architecture is sufficient on its own to address the welfare and safety challenges set out in the literature reviewed here, nor that it should displace the institutional reporting and enterprise platforms with which it would coexist. The proposition is more limited: that an open, standards-aligned peer layer would address an operational gap that current arrangements do not, and that the case for testing it empirically, at scale and under appropriate governance, is on the balance of the available evidence reasonable.

More from OnSiren Research →