On a Tuesday afternoon, a group of neighbours walks through the district with phones, notebooks, and a tape measure. They are not conducting an official audit. They note the step without a ramp at the pharmacy, the pavement that narrows beside the bus stop, and the traffic light whose push button sits too high for a child using a wheelchair. They upload a photo, describe the location, and add something the municipal plan does not know: at that hour the building's shadow makes the kerb hard to see. Their map matters because it combines observation, experience, and context.
Months later, a company launches a “frictionless” routing service. Its materials include patterns drawn from that collaborative map: marked locations, times of greatest difficulty, and comments explaining why a barrier was especially hard. The neighbourhood was never told about this use. The platform says the information was public and the data were aggregated. Yet the contributors were not donating a data reserve for any business to exploit. They were helping improve one particular way of moving through their city.
The difference is neither a technical argument about open formats nor a minor legal detail. It is a question of relationship: what does a platform understand when it receives a neighbourhood contribution, and what does it owe the people who made it possible? Posting an observation to guide another person is not the same as agreeing that it may train, segment, sell, or design a commercial product. When all those destinations are buried under one “I agree” box, consent stops describing a choice and starts to resemble a waiver nobody had a realistic chance to understand.
A map can keep, or lose, the story of its contributors#
Maps often look neutral. A red dot appears to state a simple fact: there is an obstacle here. But every dot has a history. It may have been placed by someone accompanying a parent who uses a walker; by a delivery rider who knows the slope by heart; or by a teenager who sees a car blocking the dropped kerb every morning. That experience does not make anyone the exclusive owner of a street. It does require us to acknowledge that situated knowledge did not appear by itself simply because an app included an add-marker button.
Maps often look neutral. A red dot appears to state a simple fact: there is an obstacle here.
Collective attribution makes that history visible without exposing particular people. It can say that an improvement grew out of contributions from a district community, explain which observations helped identify a pattern, and name a group when that group wishes to be named. This is not decorative credit. It lets a person choosing a route understand that human knowledge stands behind it, and it gives the community grounds to discuss whether a use still respects the purpose for which people contributed.
It also guards against a familiar fiction: that a company “discovered” barriers on its own because its system calculated them. If the calculation relied on photos, comments, or checks carried out by neighbours, hiding that origin erases labour and decision-making power. The innovation may look cleaner, but it becomes less honest. Acknowledging a contribution does not require revealing names, home addresses, or routines. It requires not presenting as spontaneous extraction what began as shared effort.
The traffic-light example reveals another difficulty. The photo may identify nobody, but a note about school pick-up time describes a local routine. Combined with other information, it could support mobility profiles or commercial decisions that were never explained. It is not enough to call one isolated item harmless when commercial value arises precisely from combining thousands of items. Privacy is not only about deleting a name; it is also about limiting which new relationships may be manufactured from a contribution made for a different purpose.
Consent is not one click followed by forgetting the journey#
Useful consent begins by separating uses. The contribution screen can first ask whether someone wants to publish an issue to guide the community and improve the civic map. Then, in a different option that is not preselected, it can explain whether there is a research proposal, a municipal collaboration, or a commercial service. Each use needs a plain-language description: who will receive the information, what they will use it for, which part of the material they need, how long they will keep it, and what they will not be allowed to do.
That separation changes the conversation. A mobility cooperative might use markers to produce accessible routes, but it should explain whether it charges for the service, shares results with other operators, or keeps original photos. A university might study areas with more obstacles, but it should explain how its report will avoid pointing to particular people. Neighbours can accept some uses, refuse others, or decide they want to discuss them first. That is participation. A blanket acceptance is barely permission administration.
The option not to agree must have an understandable effect. If someone refuses commercial use, their report must not vanish from the community map as a penalty or be relegated to a useless form. The platform should plainly explain what will continue to work and what will not. Otherwise it creates a false choice: cooperate with the business or lose a tool that may help prevent a fall. Freedom of consent is tested most clearly when saying no remains practical.
Nor should the system ask for more than is needed to mark a barrier. A picture of the kerb, an accessibility category, and a short note may be sufficient. Requesting contact details, precise phone location, journey history, or advertising identifiers because they “might improve the service” reverses the burden: it turns a civic contribution into an interrogation. If additional information is ever needed, it should be requested for a specific purpose, with an option not to provide it, and without treating silence as authorization.
Withdrawal must mean more than closing an account#
Withdrawal is the uncomfortable test of any promise of control. Imagine that, after learning about the commercial agreement, a neighbour wants to remove two photos and a comment. A hidden link labelled “manage privacy” does not solve the problem if it requires remembering dates, locating each marker, and interpreting a list of suppliers. A reasonable route lets people find their contributions, choose what to withdraw, see where it has been used, and receive readable confirmation of what happens next.
Not every consequence can be identical. A report already included anonymously in an aggregate publication may not be removable from a published statistic, while the original photo should be removable from an active catalogue. The platform needs to explain that distinction before asking for consent, not unveil it after someone objects. It should say which copies will be erased, which data remain only as a minimal record of the request, and which future uses will stop. If it cannot honour withdrawal in an external system, it should not promise total control.
Withdrawal also needs human responsibility. An automated mailbox without a deadline leaves a person facing a machine that may ignore them. The service should state who handles requests, what response time applies, and how a problem can be escalated. That person does not need to judge whether a neighbour's experience is “important enough”; they need to make sure the promised path works. When a platform gains value from thousands of small contributions, it must be able to treat the small decisions of contributors with care.
Return turns extraction into a reviewable relationship#
The final element is often forgotten: return something understandable. If a company uses community information to improve routes, the neighbourhood should be able to see what the service learned, which changes were implemented, and where limitations remain. This does not require handing over proprietary code or promising impossible results. It requires an honest return: an updated map, a list of prioritised crossings, an explanation of corrections, and a clear note about which data were not used.
That return makes quality discussable. It may show that barriers in a district are not only a matter of “reduced mobility,” but also of prams, fatigue, age, rain, limited sight, or carrying shopping. It also helps identify errors: a location may have been repaired, a photo may be outdated, or a recommended route may cross temporary works. The community stops being a store of data and becomes an interlocutor again—one that corrects, qualifies, and decides whether to continue contributing.
Before opening a new call for reports, a team can run a simple test. Bring together several people who use the map and ask them to explain, in their own words, four things: what their contributions will be used for, who may benefit commercially, how something can be withdrawn, and what they will receive in return. If their answers diverge, or depend on having read endless terms, the design is not ready.
Technology can connect scattered observations and make barriers visible when they would otherwise remain silent. It should not turn that capacity into unlimited permission to appropriate a neighbourhood's experience. The next step is to review the contribution screen: separate purposes, offer collective attribution, enable verifiable withdrawal, and publish a return commitment. Then the map can remain a tool for moving better, rather than an invisible quarry from which others extract value unseen.
