When the Tunnel Walls Started Talking: What Madrid's Bluetooth Beacons Reveal About the Future of Cities

At street level the city is still negotiable. There are landmarks: the curve of Gran Vía, the slope down to Lavapiés, the hard edge where Salamanca gives way to the M-30. But in the tunnels everything becomes a geometry exercise with too many variables. Lanes split and rejoin, exits appear from nowhere. The radio signal fades. The phone in your hand becomes a nervous animal, spinning its blue dot in circles. Space stops being a map and becomes a probability.

Then the city quietly changes the rules.

The day the tunnels started answering back

In 2025, Madrid began installing thousands of small Bluetooth beacons along the walls and cable trays of the M-30 tunnels, spacing them roughly every 20–40 meters across almost fifty kilometers of underground roadway. For around 141,000 euros, the city turned an ordinary beltway into a conversation between steel, asphalt and code.

When a driver now enters the tunnels with Bluetooth activated, apps like Google Maps and Waze no longer fall silent. They listen for these beacons, triangulate the signal, infer not only "you are in the tunnel" but "you are in this lane, approaching this exit." The navigation voice continues in its unhurried tone: "In 300 meters, keep right." Guidance persists where the satellite signal ends.

This is not spectacular technology. It is cheap hardware, simple physics and a software update. But it marks an important shift in how cities relate to the devices in our pockets. Madrid did not ask drivers to use a municipal app, or demand they buy new equipment. Instead, the city adapted its infrastructure to the tools people already use every day.

The M-30 tunnels are a laboratory of a larger story: the slow weaving together of streets, sensors, software and habits into something that we call, with varying degrees of enthusiasm or unease, "urban technology."

Cities were networks long before they were "smart"

Urban theory has been trying to catch up with this entanglement for two decades. One line of work describes cities as vast tangles of infrastructure and information, where pipes, cables and algorithms shape urban life as much as zoning codes or political speeches. Graham and Marvin's work on "splintering urbanism" argued that infrastructures no longer behave as neutral public utilities but increasingly as fragmented, unequal networks, stitching some districts tightly into high-quality services while leaving others with patchy, unreliable connections.

Another line of thought proposes that the city itself is better understood as a set of interacting systems than as a stable container for social life. To "see like a city", as Amin and Thrift put it, is to see these overlapping logics—transport, finance, logistics, platforms, families, protests—as they intersect.

Smartphone navigation is one of those logics. It is banal and universal. It reorganises how people experience distance and direction. It reroutes tourists through side streets, pushes ride-hail drivers into residential shortcuts, and turns what used to be a mental map into an external service. From a human-geography perspective, navigation apps are not just utilities; they are everyday technologies that co-produce urban space. They redraw what is accessible, what feels near or far, what is "known" or "avoid at all costs".

Madrid's beacons make that co-production explicit. The city is not simply being mapped by private platforms; it is now emitting signals designed for them. The tunnel wall becomes another line of code in the stack.

The city as a stack of layers

It helps to imagine a walk through Madrid as moving through layers.

At the bottom is the physical city: the granite curbs, the manholes with the bear and strawberry tree, the tight turns where nineteenth-century streets meet twentieth-century traffic engineering. Above that runs the layer of regulations: speed limits, low-emission zones, parking rules, bus lanes, pedestrian priority streets. Then comes the data layer: live traffic feeds, pollution sensors, transit vehicle locations, GPS traces. And on top of that, the interface layer: the apps that translate all of this into coloured lines, icons and gentle instructions.

The stack is not neutral. Each layer privileges some things and hides others. Navigation apps are very good at time and distance; they are nearly blind to atmosphere, fear, beauty or boredom. They know that a route is slower today than yesterday but not that it passes the school where you first kissed someone, or the alley you never cross after dark. They flatten topography into friction.

When the city starts to build for these apps—by installing beacons, publishing open data feeds, or negotiating API access—it is making a choice about which layers matter most. It is deciding that machine readability is a public value, that coordination between asphalt and algorithm is part of infrastructure management.

Madrid's beacons are a small but clear example of this choice. The city accepts that navigation happens in private apps, then ensures that public infrastructure speaks a language those apps understand. In this sense, the beacons are cousins to open transit APIs, air-quality dashboards or shared-bike availability feeds: ways of making the city legible to code.

The gentle violence of convenience

The story does not end with convenience. Technology in cities always carries two currencies: time saved and power redistributed.

On the surface, the beacons are about safety and efficiency. Continuous navigation inside the tunnels reduces last-second lane changes and missed exits, which in turn lowers crash risk and smoothing traffic. With around half a million vehicles a day using the underground sections of the M-30, marginal improvements in driving behaviour accumulate into fewer sirens and shorter commutes.

But there is another side. When navigation becomes perfectly reliable, people can drive deeper into the maze without fear of getting lost. The city absorbs more flows with less friction. The psychological cost of complexity falls; the environmental cost of keeping that complexity moving remains.

This is a recurring pattern in urban technology. Each wave of innovation promises to eliminate uncertainty: ride-hail platforms remove the uncertainty of hailing a taxi; food delivery apps remove the uncertainty of a kitchen and groceries; predictive policing claims to remove the uncertainty of where crime will occur. The city becomes not just a place but a series of services.

Critical work on "digital geographies" has been insisting for some time that this service-ification is not neutral. Leszczynski and others note how digital systems produce new geographies of inclusion and exclusion: who is visible in datasets, who is profiled, who is targeted with offers or enforcement.

Urban technologies like the beacons sit squarely in this tension. On one hand they are humble hardware, doing a small job well. On the other, they are pieces of a larger apparatus in which everyday life is increasingly mediated by private platforms whose business models revolve around data extraction and behavioural nudging. Sadowski's work on "smart" technologies describes how these systems are driven by a twin imperative: collect more data, exert more control.

Madrid's beacons are unusual because they break this pattern just a little: they send a one-way signal that enables navigation but do not themselves collect personal data. It is still the apps that harvest attention and behaviour, but the hardware in the tunnel is, at least technically, indifferent to who is passing by. That choice matters. It demonstrates that urban tech can support digital services without becoming another eye in the wall.

Mental maps and algorithmic maps

There is a quieter geography at work here: the interior cartography people carry in their heads.

Before satellite navigation, driving across a city required rehearsal. You would study a paper map, memorise a chain of landmarks—bridge, gas station, traffic light, billboard. Getting lost was common and socially acceptable; people asked directions at bars and petrol stations, performed complicated gestures in the air, argued about the "best way" as if there were only one.

Navigation apps short-circuit that labour. They allow residents who arrived last week to move like old hands. For people with limited time, cognitive bandwidth or physical ability, this is liberation. For those who used to live off their spatial skills—taxi drivers, couriers—it is a form of deskilling.

In Madrid, this shift is visible in how people talk about the city. Many can describe their daily routes with precision ("I take the Cercanías from Atocha to Nuevos Ministerios, then Line 10") but struggle to visualise the overall layout. The M-30 itself is often understood less as a ring in space and more as a series of on-ramps and off-ramps in the app.

When the GPS signal failed in the tunnels, this algorithmic geography broke down. Drivers dropped back into older habits: memorising exits, reading overhead signs, using senses rather than software. The beacons restore the continuity of the digital layer; they also reinforce the habit of surrendering orientation to the machine.

From a human-geography perspective, this is significant because sense of place is part memory, part practice. If the practice of navigation becomes entirely outsourced, the city risks becoming less of a lived landscape and more of a scrollable database. It is still walked, driven and cycled, but its internal coherence weakens. People know addresses rather than neighbourhoods, routes rather than districts.

The challenge, then, is not to reject navigation technologies but to design urban tech that supports rather than erodes thick relationships with place.

Madrid as a living laboratory of urban tech

Madrid's experiment with tunnel beacons is not an isolated gesture. The city has been slowly re-wiring the relationship between movement, technology and space on several fronts.

Low-emission zones and traffic‐calmed areas around the centre use license-plate recognition and sensor networks to control which vehicles can enter. Bus priority systems coordinate traffic lights to keep public transport moving. The BiciMAD bike-sharing system uses real-time data to manage dock availability and rebalance fleets across hills and valleys. Open data portals publish everything from traffic counts to air-quality readings.

Together, these projects position Madrid somewhere between the corporate "smart city" pitch and an emerging model of practical digital urbanism: build systems that solve specific problems, let them plug into the wider ecosystem, and correct course when they generate new inequalities or bottlenecks. In other words, city hall learns by iterating on the stack rather than buying a full solution off the shelf.

The tunnel beacons fit this pattern. They are a targeted response to a well-defined issue—GNSS denial in tunnels—and they rely on existing platforms rather than proprietary municipal apps. They require drivers to do something simple and voluntary (activate Bluetooth, update their app), and the value is immediately perceptible. There is no glossy branding, no utopian promises.

From the perspective of Atlas Urbium's broader concerns, this incremental approach matters. Earlier essays here have already argued that "smart cities" should be judged by how they protect encounter, ecological democracy and human-scale movement rather than by how many sensors they install.

Madrid shows one possible path: take seriously the technologies residents already use—navigation apps, messaging platforms, bikes with QR codes—and adapt infrastructure to them, while using regulation to steer behaviour towards cleaner, safer forms of mobility.

Splinters in the network: who benefits, who pays

The risk, as always, is that the benefits of urban technology flow along existing lines of privilege.

Highway navigation below ground mostly serves people who own cars. The same beacons that make the M-30 more comfortable for commuters also make it easier for ride-hailing vehicles to use the tunnels as high-speed corridors. If, as Graham and Marvin suggested, infrastructures tend to "splinter" into high-quality networks for some and decaying networks for others, urban tech can reinforce that pattern.

Workers who move through the city's digital infrastructure—couriers, delivery riders, ride-hail drivers—are also tracked, evaluated and disciplined by ranking systems and algorithmic management. They experience the city as a series of time-windows and pick-up points, not as a continuous territory. Navigation is not a convenience but a requirement. Every minute saved or lost is a number on someone else's dashboard.

In neighbourhoods where connectivity is poor, devices are older, or residents are undocumented and wary of surveillance, access to these benefits is uneven. A beacon-enabled tunnel may reduce stress for one group and do nothing for another that relies on buses or metro lines that remain overcrowded and under-prioritised.

Human geography insists that we read technologies not only for their efficiency but for their spatial justice. Where, exactly, do the gains accumulate? Who is more visible to the system, and who remains invisible? Which bodies are optimised for, and which are merely tolerated?

Madrid's broader mobility choices complicate the picture. The same city that invests in beacons is also experimenting with congestion control, dense public transport, and gradual reallocation of street space to walkers and cyclists. These moves suggest an urban project in which technology is one tool among many for shifting the city away from car dependence, not a substitute for political choices about space.

From smart cities to situated technologies

The story of urban technology over the past decade has been crowded with grand narratives. Corporate visions promised frictionless "smart cities" governed by dashboards and predictive algorithms. Critics responded with equally sweeping counter-narratives, warning that such cities would become machines for surveillance and control.

What gets lost between these poles are the situated, sometimes messy experiments in places like Madrid: Bluetooth radios screwed into tunnel walls, bike docks squeezed into old plazas, sensors taped to lampposts. These elements do not add up neatly to a masterplan. They are often compromises, partial implementations, or the material residue of negotiations between departments, vendors and community groups.

In that sense, Madrid's beacons are not a symbol of a "finished" smart city but of a city still learning how to live with its own digital reflection. The beacons solve a problem that was bothering both drivers and platforms. They do so cheaply, with limited data collection, in partnership with companies that could easily have been treated as adversaries.

This is closer to what some scholars now describe as "digital geographies" of the city: attention to how specific technologies reshape socio-spatial relations, rather than general claims about technology as salvation or doom.

Madrid's approach also resonates with a broader shift in critical urbanism away from asking "Is the city smart?" and towards "For whom, and in which places, does technology make life more liveable?" That shift opens room for questions of care, maintenance and everyday practice.

Three principles for humane urban technology

By following the beacons through the tunnels, we can outline three principles that might guide urban technology more widely—not as commandments, but as working hypotheses.

First, complement instead of replace.

The beacons do not eliminate existing forms of navigation. Street signs remain. Road markings still signal exits. Professional drivers keep their acquired knowledge of tricky interchanges. The new system overlays an additional channel without erasing older ones. In human-geography terms, this respects multiple "ways of knowing" the city: embodied, visual, digital.

Second, speak in public dialects.

Rather than demanding that residents adopt a proprietary app, Madrid chose to make the tunnels legible to tools already in use at everyday scale. This is an inversion of typical smart-city logic, where residents are asked to conform to municipal platforms. The beacons become part of a shared protocol: anyone who implements the standard can, in principle, use them.

Third, limit the appetite for data.

The beacon system is designed as a one-way signal. It allows devices to compute their position without broadcasting identity back to the infrastructure. In an era where "smart" often means "extract as much behavioural data as possible", this is a significant design choice. It shows that useful digital urbanism does not require permanent surveillance, and that privacy-preserving architectures can be deployed at scale.

These principles are not unique to Madrid. They echo in projects that use open-source air-quality sensors instead of proprietary ones, or in transit systems that expose open APIs rather than locking data behind NDAs. They point towards a mode of urban technology that is modest, interoperable and accountable: able to slot into everyday life without demanding constant attention.

Beyond the tunnel: technology as urban weather

Technology in cities is rarely the main character. It is more like the weather: ambient, changeable, sometimes violent, often taken for granted. Most days, people do not experience "the smart city"; they experience a failing ticket machine, a remarkably punctual bus, a corner where scooters accumulate, a tunnel in which the navigation voice now continues where it used to fall silent.

Madrid's beacons are one gust in that urban weather system. They are technically small and politically quiet, yet they condense many of the dilemmas that define twenty-first century urbanism: the growing entanglement of public infrastructure and private platforms, the trade-offs between convenience and control, the uneven geographies of digital access, the possibility of designing technologies that serve without dominating.

For human geographers, these beacons are not just hardware but narrative devices. They mark a point where the map begins to speak to the territory. They invite us to trace the circuits that run from a municipal procurement office, through a tunnel maintenance crew, into a smartphone chipset and back out into the driving habits of half a million people.

From the point of view of Atlas Urbium's wider project, they also connect across essays. The same city that calibrates Bluetooth signals underground must, above ground, decide how hard to push congestion controls, how far to go in reprogramming streets away from private cars, how to redistribute space for housing, how to use data to support ecological democracy rather than simply optimise flows.

Meanwhile, drivers continue to make their daily journeys. Some still memorise their exits out of habit. Others trust the calm insistence of the app. A few will never quite get used to the idea that the walls themselves are now talking, whispering coordinates into their devices.

At some point in the future, when semi-autonomous vehicles are more common and the boundary between "driver" and "passenger" blurs, the beacons may look like early fossils of a new layer in the urban stack: the layer where the city provides not only roads and rules but continuous machine-readable context for moving objects.

For now, they are simply there, emitting their silent signals in the darkness, turning a once confusing tunnel system into something more navigable. The technology disappears into routine. The city becomes, by a small degree, more hospitable to people with poor senses of direction, to newcomers, to those who do not have time for elaborate mental maps.

Urban technology at its best feels like this: not an app screaming for attention, not a glossy dashboard promising total control, but an invisible handrail in a difficult passage. You only notice it if it is missing.

Selected references

Amin, A. & Thrift, N. (2017). Seeing Like a City. Polity Press.

Batty, M. (2018). Inventing Future Cities. MIT Press.

Graham, S. & Marvin, S. (2001). Splintering Urbanism: Networked Infrastructures, Technological Mobilities and the Urban Condition. Routledge.

Greenfield, A. (2013). Against the Smart City. Verso.

Leszczynski, A. (2015). "Digital turn, digital geographies?" Progress in Human Geography, 39(3), 466–486.

Mattern, S. (2017). Code and Clay, Data and Dirt: Five Thousand Years of Urban Media. Princeton Architectural Press.

Sadowski, J. (2020). Too Smart: How Digital Capitalism is Extracting Data, Controlling Our Lives, and Taking Over the World. MIT Press.

Townsend, A. (2013). Smart Cities: Big Data, Civic Hackers, and the Quest for a New Utopia. W. W. Norton.

Nordic Semiconductor (2016). "Navigate inside tunnels with Bluetooth beacons." Get Connected Blog.

Google / Waze (n.d.). "How Waze Beacons work." Partner Support Documentation.

Madrid City Council (2025). Various press releases and media reports on the installation of Bluetooth beacons in M-30 tunnels.

Atlas Urbium. "Reclaiming the Human in the Smart City: 2025 and Beyond." Atlas Urbium – Essays.

Atlas Urbium. "Designing the Post-Congestion City: A Human-Geography View." Atlas Urbium – Essays.

Atlas Urbium. "The City That Makes Its Own Weather: How Far Can Urban Energy Go?" Atlas Urbium – Essays.

Atlas Urbium. "Urban Wounds: When Highways Become Scars." Atlas Urbium – Essays.

Atlas Urbium. "Ecological Democracy and the Big Cities: Imagining an Urban Future." Atlas Urbium – Essays.