The project links two kinds of data through the road segment. Open data describes the existing pedestrian network. Residents add proposals for where a new crossing should be. Each proposal is matched to the road network and can be linked to one or several road segments that it crosses. These associations allow comparison with crossing gaps and walking detours without mistaking a multi-road proposal for several distinct submissions.
| Data | Source | Use |
|---|---|---|
| Road Section Line, Road Crossing, Pedestrian Overhead Bridge and Underpass, Bus Stop Location, Train Station Exit (September 2026) | Land Transport Authority, LTA DataMall | Road segments, existing crossings and destinations |
| HDB Existing Building | Housing and Development Board, data.gov.sg | Residential blocks |
| Road and footpath network | OpenStreetMap contributors (ODbL) | Walking routes |
| Address search | Singapore Land Authority, OneMap API | Search box on the map page |
| Pedestrian accident statistics | Singapore Police Force annual road traffic situation reports | Context on the home page |
| Reported fatal cases | News and court reports, each linked on the home page | Context on the home page |
| Crossing proposals, experiences and support | Residents, through the proposal form | Resident proposals layer |
LTA road sections are split into segments of about 50 m. For each segment, the straight-line distance from its midpoint to the nearest road crossing, overhead bridge or underpass is measured in SVY21 (EPSG:3414). Segments are grouped into four classes, under 50 m, 50 to 100 m, 100 to 200 m and over 200 m. Expressway segments (road category CAT1) are excluded because pedestrians are not expected to cross them.
Each HDB block is paired with every bus stop and MRT station exit within 500 m in a straight line. Pairs closer than 50 m are excluded. The walking distance for each pair is the shortest path on the OpenStreetMap network, computed with Dijkstra's algorithm. A total of 227,808 pairs are used.
Detour ratio = walking distance / straight-line distance
A ratio of 1.0 means the walk is as short as the straight line. A ratio of 2.0 means the walk is twice as long. Each block is coloured by the mean ratio of its pairs (Loo et al., 2024).
A contributor marks A and B on opposite sides of a road, at least 2 m and no more than 80 m apart, and selects a crossing type and main reason. The new backend checks which road centrelines the marked crossing intersects. A single proposal can cross two roads or separate carriageways, so it is stored once and linked to each associated road segment. Only validated, accepted proposals appear in public statistics; new submissions require review, while eligible imported pilot records with status ok may count as previously accepted when their road IDs remain compatible. A road can appear from its first accepted proposal. Each published road shows its count, most requested type, most common reason and distance to the nearest existing crossing. One multi-road proposal contributes once to the overall proposal total but may contribute to each associated road count. Per-road counts must therefore not be added together to calculate the total number of proposals. The displayed crossing line is rounded or aggregated; if based on one accepted proposal, its approximate selected location may be visible rather than a multi-person average.
Visitors can also add support to a segment that is already shown. Support is counted separately from proposals.
Processed data used on the map, in WGS84 (EPSG:4326).
Loo, B. P. Y., Lian, T., & Frank, L. D. (2024). Walking (in)convenience: An in-depth study of pedestrian detours to daily facilities. Journal of the American Planning Association, 90(4), 742–757.