From Vision to Formula: Turning Famous Urban Planning Rules Into Measurable Accessibility Maps

Every few years, a planning slogan goes viral: the 15-minute city, the 3-30-300 rule, the idea that walking should be an integral part of every public transport journey. These numerically expressed heuristics condense sprawling sustainability ambitions into memorable propositions that planners, politicians and residents can actually discuss. But a slogan is not a map. The distance between a viral vision and a reproducible spatial benchmark is where most of the methodological difficulty hides, and a new study in Frontiers in Sustainability has now built a formal bridge across it.

Researchers Giannis Adamos of Aristotle University of Thessaloniki, Tamara Rađenović of the University of Niš and Milena Rajić of the University of Niš have developed the Numerical Urban Sustainability Integration Framework, or NUSIF, a spatial-accessibility framework that converts selected operational dimensions of three distinct planning traditions into comparable spatial evidence. Rather than claiming to test full compliance with any of the underlying concepts, the framework extracts three spatially compatible dimensions: accessibility to essential urban services, inspired by the proximity logic of the 15-minute city; accessibility to public green space, derived from the 300-meter proximity component of the 3-30-300 rule; and pedestrian access to public transport, motivated by the walking-public transport interface associated with the 90-50-70 practitioner heuristic.

The technical logic is deliberately transparent. The framework was applied to Niš, Serbia, using pedestrian-network distance rather than straight-line buffers, a nominal 500 by 500 meter analytical grid, OpenStreetMap-derived destinations and network data, and a modeled 2025 population surface from the European Commission’s Global Human Settlement Layer. After quality control, the final analytical dataset comprised 873 unique cells representing approximately 287,673 residents. Essential-service accessibility was evaluated for six functional categories, covering food, health, education, daily services, culture and recreation, within 1,200 meters of pedestrian-network distance. Green-space accessibility was assessed at 300 meters, and public-transport accessibility at 500 meters. Each component was normalized to a common zero-to-one scale and combined using an equal-weight arithmetic specification, producing a composite score that was then divided into five fixed interpretation classes.

The baseline results reveal a strongly differentiated accessibility structure. The unweighted cell mean NUSIF was 0.2365, with a median of just 0.1111. Under the fixed classification, 492 cells, or 56.36 percent of the analytical territory, belonged to the lowest NUSIF class, and a further 183 cells were classified as low, meaning that 77.32 percent of all cells fell into the two lowest categories. Only 96 cells, about 11 percent, reached the high or very high classes. On purely territorial terms, Niš appears as a city where very low integrated accessibility is the dominant spatial condition.

But the population picture is strikingly different. The very-low and low classes contained only 10.93 and 11.49 percent of the modeled population respectively, while the high and very high classes, although comprising just 11.00 percent of all analytical cells, concentrated 61.79 percent of the modeled population. Roughly 103,920 residents lived in high-NUSIF cells and 73,846 in very high cells. This divergence between territorial coverage and population exposure is one of the central empirical findings of the study: spatial disadvantage and population disadvantage are different analytical questions, and a single citywide indicator cannot adequately represent both.

To move from diagnosis to intervention analysis, the researchers extended NUSIF through three independent counterfactual scenarios, each modifying one component while holding the others constant. Scenario 1 added one accessible service category to every cell not already attaining the maximum of six, Scenario 2 expanded the effective green-space network threshold from 300 to 500 meters, and Scenario 3 expanded the effective public-transport catchment from 500 to 800 meters. Importantly, the scenarios perturb analytical accessibility conditions rather than modeling new infrastructure, and because they modify differently scaled components, they are interpreted as diagnostic counterfactuals rather than cost-equivalent policy alternatives.

The scenario effects were strongly criterion-dependent. Scenario 1 produced by far the widest spatial reach, changing the score in 810 cells, or 92.78 percent of the territory, containing about 55.34 percent of the modeled population, and raising the population share in the two highest classes to 68.96 percent. Yet because the maximum cell-level increment was comparatively small, only 128 cells crossed a classification threshold. Scenario 2 changed the score in just 22 cells, but every affected cell crossed at least one class boundary, and the population represented by the very high class surged from 25.67 to 41.38 percent, because those 22 cells contained roughly 16.68 percent of the study population. Scenario 3 produced the broadest territorial redistribution, reducing the number of very low cells from 492 to 402, the largest reduction among the three scenarios, while affecting cells containing only 7.66 percent of residents.

The ranking reversal between evaluation perspectives is perhaps the most consequential result. Based on the unweighted cell-mean change, the scenarios ranked S1 greater than S3 greater than S2. When the distribution of the resident population was taken into account, the ranking became S2 greater than S1 greater than S3. The population-weighted baseline NUSIF of 0.6106 was also substantially higher than the unweighted cell mean of 0.2365, reflecting the concentration of residents in comparatively accessible cells. This means that intervention appraisal based on a single aggregate score can be genuinely misleading: the scenario producing the broadest spatial improvement was not the scenario producing the greatest population-weighted gain.

The researchers also tested the robustness of these findings through a weighting sensitivity analysis across seven alternative component-weight specifications, ranging from moderate to stronger emphasis on each of the three components. The baseline unweighted cell mean ranged from 0.1565 to 0.2803 under alternative weights, and the population-weighted value from 0.4752 to 0.6854. Yet under every specification, the population-weighted baseline remained substantially higher than the corresponding unweighted cell mean, and cell-level rank ordering remained highly stable, with Spearman rank correlations between alternative-weight and equal-weight configurations ranging from 0.9655 to 1.0000. The central distinction between territorial and population-weighted evaluation therefore persisted across all seven weighting structures, although the precise ordering of individual scenarios proved weight-sensitive, changing when greater importance was assigned to the component directly affected by a given intervention.

The authors are careful about what NUSIF does not claim. The service component captures functional diversity rather than facility abundance, the green-space component measures network proximity but not tree visibility, canopy coverage or environmental quality, and the public-transport component represents first and last-mile spatial availability without service frequency, reliability, fares or walking-environment quality. The framework also relies on OpenStreetMap data, whose completeness can vary, and the 500 by 500 meter grid introduces the well-known modifiable areal unit problem. What NUSIF offers instead is a transparent integration architecture through which established but conceptually distinct accessibility dimensions can be examined jointly, preserving component-level information, distinguishing territorial from population-weighted outcomes, and comparing component-specific counterfactual response profiles. For planners, it is best understood as a screening and diagnostic tool, one that identifies where accessibility improvements occur, how substantial they are, and how strongly they coincide with the spatial distribution of residents, while implementation decisions still require costs, feasibility, service capacity and community priorities to be weighed. In an era when planning slogans travel faster than evidence, that kind of structured translation from vision to formula may be exactly what sustainable mobility needs.

Subject of Research: Urban Sustainability

Article Title: From vision to formula: integrating numerical urban planning heuristics into a spatial accessibility framework for sustainable mobility

Article References: Adamos, G., Rađenović, T., & Rajić, M. (2026). From vision to formula: integrating numerical urban planning heuristics into a spatial accessibility framework for sustainable mobility. Frontiers in Sustainability, 7, Article 1971777. https://doi.org/10.3389/frsus.2026.1971777

Image Credits: AI Generated

DOI: 10.3389/frsus.2026.1971777

Keywords: urban accessibility, 15-minute city, 3-30-300 rule, sustainable mobility, GIS, population-weighted accessibility, urban planning, NUSIF, vision, formula, integrating, numerical