Observing a city block: a reference design

01What a city block needs to see

Observing a city block starts with the air people breathe. Cities report on air quality as part of the sustainable development goals, and one indicator is the annual mean level of fine particulate matter in cities, weighted by population¹.

Health is the reason. The global air quality guidelines of the World Health Organization cover particulate matter, ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide, and they find that air pollution affects health at lower concentrations than previously understood².

Heat is the second need. Streets, courtyards and parks on the same block can differ sharply in temperature, and the international guide to instruments and methods of observation gives urban observations a chapter of their own³.

Movement is the third. Footfall on a street during an event, a heatwave or an evacuation tells the city where people are. It is the most sensitive of the three, so this design counts people without identifying them.

02Siting sensors on streets and poles

Street poles are the natural home for city sensors. They are already powered, spread along every street and owned by the city. In this design each small cell pole also carries sensors, and air-quality nodes sit on lower street furniture near the height at which people breathe.

Low-cost air sensors can drift in service, so they are tested before use and corrected against a reference while they run. National guidance sets out testing protocols, performance metrics and target values for sensors used in supplemental and informational monitoring, with field and laboratory testing⁴.

This design places one air-quality node beside a reference station and uses it to derive the correction for the others. The correction, and the date it was made, travel with every corrected value.

Heat sensors need shade and ventilation, or they measure the sun on their own housing. The international guide has chapters on measuring temperature and on urban observations³.

Small cells sit at pole height and serve the street below. Propagation methods for planning short-range outdoor radio systems from 300 MHz to 300 GHz are set out in an international recommendation⁵.

03From sensor to decision

Sensors on the street send short reports to the nearest small cell. How these base stations transmit and receive is set in a mobile network specification for base station radio⁶.

The small cells connect to an edge node in the district. Edge computing standards define an open environment in which applications from different providers can run on multi-vendor edge platforms⁷. Here the edge node applies the air-quality correction, checks every value and turns raw counts into totals.

Two kinds of user read the result. City operations see live values for traffic, events and heat warnings. Everyone else sees open data, published through an open interface for sensor observations and their metadata⁸.

Privacy is designed in, not added later. Footfall leaves the edge only as totals per street and interval, never as images, device identifiers or tracks.

Each value that leaves the edge carries its context: the time, the place, the sensor, whether it was corrected and how, and the quality checks it passed. A planner reading a heat map next summer can then tell a measured street from an estimated one.

04Density and spectrum are the constraint

In a city, the hard limit is not range but density. Buildings block and reflect the signal, so a small cell serves a short stretch of street, and many sensors share each cell and its spectrum.

The planning trade is between cell radius, the number of sensors and how often they report. Fewer, larger cells mean more sensors per cell; frequent reports mean more data on shared spectrum. Planning that radius uses the propagation methods for short-range outdoor radio systems⁵.

Interference is the other side of density. Small cells on neighbouring poles reuse the same spectrum, so their power and placement are planned together, and sensors that report on a fixed schedule are spread in time so that they do not all speak at once.

The simulator shows the trade for a small district. Change the cell radius and see how many sites are needed, then change the report interval and see how the daily data volume follows.

05When the links fail

Streets are busy places, and poles get hit, dug up and switched off. Every sensor in this design can reach a second small cell, so the loss of one pole moves its sensors to the next rather than taking them offline.

When the link from the district to the city is lost, the edge node keeps checking, correcting and storing. City operations on the local network still see live values, and open data catches up in order, with original timestamps, when the link returns.

A sensor that fails, or that drifts out of its correction, is flagged and shown as a gap with its reason. Nothing is filled in silently, because a street with no reading must not look like a street with clean air.