The map and route review show the reports available under these rules. Missing or imprecise reports may leave gaps, so an empty result cannot establish that a route is safe.
How the pedestrian map works
Find a place scopes reports to the resulting map area. Search this area applies the current viewport; report text and map-availability filters continue to apply. List-only records stay in the scoped list because their location cannot establish whether they lie inside the area. An empty list means no matching reports in the available archive, not an absence of accidents.
How route matching works
- Route construction: Pauserby joins OneMap walking legs between your selected places, including up to two places to travel via. Adding an inspection stop later does not change the route.
- Distance thresholds: Mapped pedestrian cases are included within 100 metres of the route, LTA incidents within 150 metres, and approved community reports whose selected place is within 250 metres of the route.
- Approximate locations: For pedestrian cases mapped as approximate areas, Pauserby subtracts the documented uncertainty radius from the distance to the route. Areas with a radius over 500 metres are not matched automatically. Road-only and unmapped cases remain in the full archive.
- Historical LTA incidents: The route review includes incidents observed in the past 12 months, starting from when Pauserby began collecting. It shows up to the 200 closest previously observed accidents and marks the count if more match. Current incidents have no such limit. A nearby report does not establish that an accident happened on the route.
- Evidence order: Results sort by overlap, effective distance, then recency. Numbered walkthrough stops follow the route and serve as places to inspect the evidence, since reported locations may be approximate.
Crossings and covered linkways
- Data source: The walking infrastructure layers use LTA's Road Crossing and Covered Linkway datasets. Each layer links to its source download and shows the dataset month. These are published map snapshots, not live reports of access or condition.
- Map browsing: Enabled layers load only at neighbourhood scale (zoom 14 or closer), within a viewport no wider or taller than 0.06 degrees. Each layer returns up to 500 features and 400 KB of geometry. The control identifies incomplete results and independent layer failures; city-scale browsing does not load the full snapshots.
- Route matching: Pauserby finds mapped features within 15 metres of the walking route and lists them in route order. Each layer shows up to 100 matching features and indicates when more matched.
- What the totals mean: A feature may be one part of a facility, so the totals are not counts of distinct crossings or walkways. An overlap on the map does not establish that the route uses the feature or is on the same level.
- Known gaps: Crossing data does not specify signal timing, crossing type, or accessibility. Covered linkway data may omit other shelter, including building passages and void decks. A missing feature does not establish that a path is uncovered or has no crossing. These layers do not provide shelter percentages, step-free guarantees, or route safety ratings.
Map data: Land Transport Authority. See the LTA DataMall catalogue and Singapore Open Data Licence.
Pedestrian accident archive
The archive draws on public Singapore Police Force releases and news reports of pedestrian collisions, including cases where the injury outcome is unknown. Each case links to its sources and records the wording behind its map location. Some reports identify only a road or give too little detail to map; those cases stay in the list. Public reporting is incomplete, so the archive is not a full accident history.
Community reports
Verified members can submit a close shave or accident with a date, description, and nearby public place. A moderator checks each submission for clarity and privacy before publication. The place is an approximate reference, and the account has not been independently verified as an accident.
LTA traffic accidents
- Collection: Pauserby checks LTA's Traffic Incidents feed every two minutes and keeps each response and observation. A poll affects incident status only when every page was received and decoded.
- Incident grouping: Because the feed has no documented stable incident ID, Pauserby groups observations that appear to describe the same event. An incident ends after three consecutive successful polls without a matching observation, and a strong match can reopen it within 15 minutes.
- Limitations: LTA coordinates are approximate. The feed does not state whether pedestrians were involved or provide casualty and severity details. Pauserby's history covers only incidents it has observed, while the Traffic accidents page also shows what the feed currently reports.