ALX Dispatch.

The data

Everything on the map comes from one file. This page is for anyone — a reporter, an intern, NVFSS, a civic technologist — who wants to use the dataset directly.

Download

incidents.csv — one row per incident, cumulative since April 2026. Plain CSV, updated monthly.

The map fetches this exact file at load time, so what you download is what you see.

What this dataset is

A corridor incident tracker: every pedestrian, cyclist, and scooter strike captured citywide, plus vehicle-only crashes and other incidents on the two advocacy corridors — Braddock Road and Mount Vernon Avenue. The two are tracked together because corridor crash density is the evidentiary argument for the next Braddock fight (Phase 2 / Braddock West). A single row can be both a VRU strike and a corridor incident; the vru_strike and corridor fields are independent flags, so every view on the map falls out of filtering.

Column dictionary

ColumnMeaning
dateISO date (YYYY-MM-DD).
time24-hour HH:MM, local.
dayDay-of-week abbreviation.
areaNeighborhood / area label.
locationIntersection or address string.
lat, lngApproximate intersection/corridor-level coordinates (see geo_precision).
geo_precisionintersection / address / block / approximate.
corridorBraddock / Mount Vernon / blank. Marks advocacy-corridor rows.
phaseBraddock only. 1 = east (Better Braddock / Phase 1); 2 = west (Minnie Howard area / Braddock West).
vru_strikeY = a pedestrian/cyclist/scooter/motorcycle was struck; N = vehicle-only / other. Drives the dot color.
modepedestrian / cyclist / scooter / motorcycle / vehicle / DASH bus / police / unknown.
incident_typestrike / fatality / vehicle_crash / hit_and_run / police_vehicle / of_interest. Kept separate from mode — a two-car crash is never coded as a VRU strike.
hit_and_runY / N.
drca_bordersY if within DRCA boundaries: Glebe (N), Russell (E), Braddock (S), Route 1 + CSX tracks (W).
reportedY if it received news / APD-social coverage.
severity_noteFree text: context, confirmed police detail, advocacy notes.
dispatch_summaryOne-line paraphrase of the dispatch audio.
follow_up_askWhat we want APD to confirm (drives the monthly follow-up list).
confirm_statusopen (dispatch-sourced only) / confirmed-context (APD/news detail obtained).
fr300p_filedy / no / unknown — filled from APD follow-up.
citationY / no / unknown — filled from APD follow-up.
source_monthApril / May / …
incident_idInternal id, e.g. apr-64, may-13.
openmhz_or_pageSource reference: the published month page (or OpenMHz) the incident is drawn from.
card_urlDeep link to the published card with audio, e.g. /archive/2026-05/#incident-01.
child_victimY if the victim was under 18. Renders a mid-size dot and a badge.

Common views (all are filters, not separate files)

Reuse

The dataset is public and free to reuse, including by reporters and other advocates. Two requests: attribute it to this project and link back to alxdispatch.org, and carry the approximation and scope caveats below so the numbers aren't oversold. The underlying tool is open source (MIT) at dispatch-pedestrian-monitor.

Scope & caveats. VRU (pedestrian/cyclist/scooter) strikes are complete citywide from April 2026 onward. Systematic vehicle-only crash tracking on the corridors begins May 2026 — April vehicle-only corridor crashes are not comprehensively captured. Coordinates are intersection-level approximations, not exact points of impact (see geo_precision). Counts reflect what the tool caught and a human verified on the two primary dispatch talkgroups; the true totals are almost certainly higher. See the Methodology for the full picture.