April 1–22, 2026 · Alexandria, Virginia · A personal research project
Eighteen people were hit by drivers in Alexandria in twenty-two days.
Six made the news. One pedestrian died. Every incident below is documented with the original dispatch audio.
18
Total strikes
6
Publicly reported
7
Hit-and-runs
1
Fatalities
Alexandria has a data gap when it comes to pedestrian, cyclist, and scooter
safety, and I noticed it by accident.
This isn't a story about our police department falling short. The officers
responding to these incidents do great work. But their public reporting
role seems to be narrowed to “a road is closed” followed by
“a road is reopened” — which is why we hear about some
of these incidents on social media and not about others. Other local
cities have decided to make more of this data, and Alexandria should too.
More on that at the bottom.
I serve on the Del Ray Citizens Association's traffic calming committee,
and I attend meetings for the Bicycle and Pedestrian Advisory Committee
(BPAC) and Northern Virginia Families for Safe Streets (NVFSS). In each of
those rooms, there's a police representative gamely trying to reconstruct
the last month's crashes from whatever notes they have easy access to.
I've always known these presentations have gaps, because every month
incidents we already knew about — from text threads that run parallel
to everything in Del Ray — would turn up missing from the official
list. That's how I learned that a crash often leaves no public record
unless there's enough property or human damage, and that even the ones
that do don't show up in the state's TREDS database for months. The gap
between the partial pictures advocacy groups get, and what the city itself
quietly hears over the radio, is enormous.
If our Vision is truly Zero deaths, we've got to close this gap.
Hearing a hit-and-run call come over the scanner at the Braddock Road
Metro last Wednesday night made me wonder how often this was happening.
So I built a tool to download every OpenMHz dispatch call that was publicly available,
transcribe them all, and search for anything that looked like it might be a
pedestrian, scooter, or cyclist incident. I reviewed every flag myself.
Below is every verified incident I could find from April 1 through April 22,
2026. Six of them had made the news. Twelve had not.
Background: what the public can see and what it can't
What other outlets reported.
Two Alexandria news outlets began connecting the dots after the
pedestrian death April 20th. The Alexandria Brief noted in an April 23
article that
“the five incidents described in this article reflect those APD
announced publicly on social media — typically because they resulted
in road closures or sustained police activity — and may not represent
every pedestrian-involved crash in Alexandria during that period.”
ALXnow published a similar piece the same day. The caveat is correct. What
follows below is an attempt to measure the gap between what APD posts and
what dispatch heard.
Alexandria's
Vision Zero dashboard
draws from crash reports that officers file with the state. Under Virginia
Code § 46.2-373, those reports are required only when there's a death,
a confirmed injury, or property damage of $1,500 or more. (For drivers in
hit-and-run collisions, a separate provision requires self-reporting at
$500 of damage.) The dashboard itself notes that data may not appear for
at least two months after a crash. At the time of this writing, no
Alexandria 2026 data has been published to the dashboard yet.
Why dispatch data is different.
Three filters sit between what happens and what the public sees: the
statutory reporting threshold, what can be confirmed on scene, and the
months-long publication lag. Dispatch audio is upstream of all three. It
captures what someone called 911 about, before an officer decides whether
a crash report is required, before the victim says they're fine and leaves,
and long before anything enters the state database.
Dispatch over-counts reality — some calls turn out minor or get
corrected on investigation. The reportable-crash dashboard under-counts
reality. Truth sits between them. Right now, the public only has access to
the under-count, on a two-month delay.
The eighteen pedestrians hit — so far in April
In chronological order. Click any clip to play.
Saturday, April 4 · 9:22 AM · Del Ray
Mount Vernon Ave & E Luray Ave
Pedestrian
Pedestrian hit at Mount Vernon and Luray at 9am on a Saturday morning — right in the middle of the Farmers Market / Matt & Tony's brunch rush. The DRCA's police report the next week on this one was "we know a vehicle incident happened at Luray Saturday morning but that's all the detail we have." Reviewing the calls, we now know it was a pedestrian struck — but because the pedestrian refused medical attention, nothing further will be recorded in any database.
Pedestrian struck Mount Vernon and East Luray, Mount Vernon and East Luray. Caller was hit by a Honda Accord, that's still on scene. Medics refused, so routine response.
9:22:03 AM · AFD · 19s
9:22:22 AM · APD · 16s
Saturday, April 4 · 4:01 PM · Del Ray / Braddock Project Area
Mount Vernon Ave at E Braddock Rd (in the Braddock Project Area)
PedestrianReportedSchool Zone
An adult male pedestrian was hit by a driver while in the crosswalk and received medical attention. We know from news coverage that the driver was cited at the time of response, which tells us there were either additional witnesses or video to confidently show the driver was at fault. Even after several conversations with officers, citizens still don't know with certainty which direction the driver was traveling from.
235 Alpha unit to assist with the crash. Pedestrian struck Mount Vernon East Braddock. Information is still coming in. Priority response is authorized.
Previously unknown before this analysis. A US Park Police officer called in a female pedestrian who had been struck by a driver.
Units 214, 224, crash 5 and 220 for the pedestrian struck South Royal and Green Street, South Royal and Green. Call came in to us from U.S. Park Police advising that they're out with a female that was struck by a vehicle. Priority response is authorized at 1220.
12:20:29 PM · APD · 19s
12:20:54 PM · APD · 26s
Tuesday, April 7 · 3:22 PM
Sanger Ave & N Van Dorn St
PedestrianReported
A serious pedestrian strike with injury at Sanger and Van Dorn. APD tweeted a street closure, which is probably the only reason this one made the news. Community meetings since have filled in two structural contributing factors: the sidewalk near the crash was closed at the time, forcing the pedestrian into the street, and a crosswalk sign in the area had been reported broken. Both are city-controlled conditions.
Female on a scooter hit by a driver. Caller said the driver was trying to leave the scene.
Headquarters any units a break for the MVC. Pit, Pedestrian struck, Seminary and Mark Center. Seminary and Mark Center. Priority responses authorized at 20-16 hours.
8:15:55 PM · AFD · 19s
8:16:44 PM · APD · 11s
8:17:19 PM · APD · 6s
8:17:42 PM · APD · 25s
Thursday, April 9 · 2:42 PM
4368 King St (Liberty Gas Station)
PedestrianHit and run
Pedestrian caller(s) - multiple - saying they were on foot and struck by a female driver in a red sedan who then left the scene. They asked for the gas station cameras to be reviewed. Because there were no injuries or property damage, this one will not be recorded in Vision Zero database.
MVC hit-and-run, no injury, pedestrian struck, 4368 King Street, 4368 King Street. Callers stating they were on foot and struck by a vehicle. Described it as a red sedan, medics have been refused so far.
2:42:47 PM · APD · 25s
Saturday, April 11 · 7:22 PM
Valley Forge Dr & S Pickett St
PedestrianReported
In this incident, AFD was responding to a residential building fire in the area, and as they did so a private vehicle that was pulling out of their way struck a pedestrian. That pedestrian required medical attention and a secondary emergency event. AlxNow appears to have caught this one on the scanner and asked APD for further comment but never published any update.
We're heading to the we're heading to the box alarm in car tried to get out of our way and hit a pedestrian in the street at Valley Forge in South Pickett. He's laying in the street.
A scooter was struck by a driver on Stovall near Mill Rd. (near the Wegmans). Because medical attention was refused on scene, this incident is not Vision Zero database reportable. No further information is available regarding whether the driver stayed on scene, but 911 call records or police records should show this.
X-ray 908. 908. Can you start a unit reference accident, hit and run, a pedestrian was struck on the street. Um, that's Stovall right before Mill.
9:07:31 PM · APD · 17s
9:08:29 PM · APD · 11s
Tuesday, April 14 · 9:32 AM
906 Wolfe St
PedestrianHit and run
Not much information here but it happened at 9:30am in Old Town along Richmond Highway, and is described as a hit and run.
213 at the back for the hit and run pedestrian struck 906 Wolfe, 906 Wolfe for a pedestrian that was struck by a vehicle. Unknown description of vehicle. Pedestrians refusing to give any further.
9:32:37 AM · APD · 22s
Wednesday, April 15 · 6:05 PM
5750 Sanger Ave (near William Ramsay Elementary)
PedestrianReportedSchool Zone
Dispatch describes this as a 7-year-old male hit by a driver; news accounts say he was 11. From a Police presentation to Northern Virginia Families for Safe Streets, I understand that this child may have been walking behind a bus at the time.
332-32-November-1, 32-November-2 for the pedestrian struck 5750 Sanger Avenue, 5750 Sanger Avenue. Priority response is authorized, 1806. You're responding for a 7-year-old male that was hit by a gray vehicle, Virginia tag, ********. We do have fire enroute as well.
Pedestrian struck in the crosswalk in Old Town during morning hours. The clips here show APD requesting a motor unit to close the intersection and being told one isn't available — which may be why this incident never made it onto APD's Twitter.
223 and any unit that could back for Franklin and South Alfred, Franklin and South Alfred for a pedestrian struck. It's going to be a gentleman that was in the crosswalk that got hit by a vehicle. The vehicle is a grey Dodge Ram stuck to the pedestrian. No visible bleeding at this time. Medics are also responding to the location. Priority response is authorized.
8:48:55 AM · AFD · 18s
8:49:12 AM · APD · 25s
8:49:43 AM · APD · 6s
8:49:55 AM · APD · 4s
8:50:05 AM · APD · 7s
8:50:29 AM · APD · 4s
8:58:14 AM · APD · 11s
Thursday, April 16 · 4:46 PM
N Beauregard St & Highview Ln
CyclistHit and run
A bicyclist was struck from behind and the driver fled — but not before handing the victim a hundred-dollar bill. Reported ~10 minutes after the fact from a bus stop near North Beauregard and Highview. Driver described as operating a white Toyota Prius.
Motor vehicle accident with injuries involving a bike occurred about 10 minutes ago at the intersection of North Beauregard and North Highview. Caller advised his rear tire was hit by a vehicle, complaining of leg pain — but the other driver gave him a hundred dollars and left. The striking vehicle is going to be a white Toyota Prius.
4:46:04 PM · AFD · 19s
4:47:07 PM · APD · 8s
4:48:58 PM · APD · 35s
Friday, April 17 · 12:09 AM
Potomac Ave & Richmond Hwy
PedestrianHit and run
Hit and run in which a female scooter rider was struck just after midnight. Responding units never get a vehicle description beyond "unknown vehicle" — the driver was gone before anyone arrived.
Hit and run, Richmond and Potomac, Richmond Highway, Potomac Avenue, priority response triple zero nine. Call is still collecting further. A female was on a scooter struck by an unknown vehicle. Fire is en route.
12:09:16 AM · AFD · 19s
12:09:27 AM · APD · 15s
12:10:05 AM · APD · 7s
Saturday, April 18 · 1:37 PM
Commonwealth Ave & Uhler Ave
PedestrianHit and runSchool Zone
Appears to have been a hit and run of a scooter rider with injury reported Saturday afternoon right across from Mount Vernon Community School.
Rescue 209, Medic 209, EMS 201, operate on 2 Bravo, Fox, 202-33, Bicyclist Struck, Commonwealth Avenue, and West Uhler Avenue. Rescue 209, Medic 209, EMS 201.
1:37:54 PM · APD · 17s
1:37:55 PM · AFD · 20s
Saturday, April 18 · 9:11 PM
Reading Ave & N Beauregard St
Pedestrian
An adult male pedestrian "struck" — though dispatch states the subject's feet were run over by a Jeep Wrangler whose driver remained on scene. (Content note: the audio is distressing.)
331, it's going to be a representing MVC pedestrian, stuck reading at North Beauregard. The driver is still on scene. It's going to be a Jeep Wrangler. Stated the subject's feet have been run over. The male is conscious and breathing.
9:11:48 PM · APD · 14s
9:11:50 PM · AFD · 19s
9:12:42 PM · APD · 14s
Monday, April 20 · 5:26 PM
N St Asaph St & Montgomery St
PedestrianFatalityReported
A 62-year-old female pedestrian was killed by the driver of a black Suburban during evening rush hour. It was originally reported as a hit and run; dispatch indicates the driver stopped at Montgomery and Washington a minute or two later.
Caller advised the striking vehicle is a black Suburban. Pedestrian is down on the ground, not breathing at this time.
5:26:58 PM · APD · 8s
5:27:02 PM · AFD · 19s
5:27:20 PM · APD · 31s
5:27:55 PM · AFD · 2s
5:28:19 PM · APD · 18s
5:29:19 PM · APD · 10s
5:29:27 PM · AFD · 20s
5:30:25 PM · APD · 6s
5:30:41 PM · APD · 7s
Wednesday, April 22 · 4:38 PM · Braddock Metro
500 E Braddock Rd at Braddock Metro ("Braddock Metro Crosswalk")
PedestrianHit and runSchool Zone
The call that inspired this project. A female pedestrian was struck in the crosswalk by a driver who then turned right onto West and was subsequently chased onto Duke St. Because traffic was not shut down and the pedestrian is ok, this one may not be reportable elsewhere, but as the second pedestrian struck this month in the Braddock Road project area, it should clearly be of public interest for council.
324 and 332 for the MVC pedestrian struck, 500 East Braddock Road, 500 East Braddock Road, priority response is authorized at 1639 hours, striking vehicle is going to be a black SUV, unknown tag, attempting to get the direction of travel. It's on the crosswalk to the Braddock Road Metro Station.
4:38:19 PM · AFD · 40s
4:38:54 PM · APD · 18s
4:39:16 PM · APD · 3s
4:39:26 PM · APD · 11s
4:39:54 PM · AFD · 2s
4:40:18 PM · APD · 7s
4:40:31 PM · APD · 10s
4:41:02 PM · APD · 6s
4:41:26 PM · APD · 7s
4:41:40 PM · APD · 12s
4:42:38 PM · AFD · 14s
4:43:55 PM · APD · 14s
4:44:16 PM · APD · 5s
4:44:50 PM · APD · 11s
4:45:18 PM · APD · 10s
4:45:50 PM · AFD · 17s
4:45:59 PM · APD · 6s
4:46:11 PM · APD · 4s
4:46:32 PM · APD · 7s
4:46:51 PM · APD · 23s
Wednesday, April 22 · 11:00 PM
King St & Park Rd
PedestrianReported
Although the original call described it as a hit and run, the driver stayed on scene for this Rosemont pedestrian strike. The call led to an APD traffic shutdown tweet, and kicked off the "this seems to be happening a lot" wave of local media coverage.
Units 112, 312, and 315, it's going to be for the accident hit and run pedestrian struck at the intersection of King Street and Park Road, King Street and Park Road. Scan priority response is authorized. We do have medics responding, again there is a victim in the road.
11:00:36 PM · AFD · 14s
11:00:46 PM · APD · 26s
11:01:18 PM · APD · 4s
What Alexandria should do
What we need isn't one fix. It's three layers, because each one catches
what the others miss — and right now, the city could be better-using
all of them.
1. Adopt a Charlottesville-style general order.
Every Virginia city operates under the same statutory reporting threshold
as Alexandria. But not every city stops there.
In Charlottesville, Police Department General Order 542.01 has been on
the books in some form since at least 1999, most recently revised in 2020.
It is a routine internal departmental policy, not a reaction to any single
tragedy:
Whether on public or private property an incident report is required for
accidents that result in death, critical injury, pedestrian/bicyclist,
Hit and Run and all Hazardous Material.
— Charlottesville Police Department, General Order 542.01
(section IV.D.1).
View the PDF.
The order creates an internal record for every pedestrian or bicyclist
incident, regardless of whether the state's reporting threshold is met.
Alexandria does not appear to have an equivalent policy. Adopting one
— and explicitly extending it to scooters, which didn't exist in
meaningful numbers when Charlottesville's order was last revised —
would close the documentation gap for the strikes officers responded to but
the state never recorded. This is a simple ask that I believe would just
require a policy change to implement.
2. Systematically screen dispatch audio every month.
The eighteen incidents documented above were found by listening to
publicly-available dispatch audio for twenty-two days, transcribing it, and
reviewing what came up. I built the tool myself, and ran it for a month
for the total cost of $21 in transcription credits, plus the AI
subscription I was already paying for, and a few hours of my time.
We have a very rich and nearly free data source sitting out there in
these dispatch records, just waiting for us to use it. This ask is that we
use it. The output isn't perfect, and we shouldn't present it as such. But
as a starting list it lets advocacy groups, community associations, and
city staff fill in context from a template rather than from a blank sheet.
It catches incidents that would otherwise disappear into refused-medical,
no-property-damage, no-paperwork silence.
The point here is definitely not to react to each individual incident.
T&ES has tried that approach in the past and rightly stepped back from
it when the crashes outpaced the engineering capacity to address them one
at a time. Instead, we should use this data to spot patterns where
the official crash record alone is too sparse to reveal them. This is a
feeder for trend analysis, not a backlog of individual repairs.
This data could play an especially critical role as we look to update
our Vision Zero Action Plan later this year. T&ES has signaled interest
in building a citywide risk-assessment framework — an objective tool
that combines inputs like vehicle speeds, lane counts, traffic volumes, and
corridor design with crash history, so a corridor's risk profile doesn't
depend solely on whether a fatal crash has already happened there.
Systematically-coded dispatch data is well-suited as one of several inputs
to a framework like that. This piece is offered as input to that effort.
T&ES staff are already spread thin, but NVFSS is positioned to
continue staffing this project going forward and I am at their service to
help make that happen. Long term, we should find a path to add verified
events with full context to existing Vision Zero reporting. The ask of
T&ES is not to do the work. It's simply to use the output.
3. Expand the DASH-NVFSS near-miss reporting partnership across city signage.
Even with both a documentation policy and dispatch monitoring in place,
two large categories of data are still missing: close calls in which no one
was hit, and incidents involving people who don't call the police.
Northern Virginia Families for Safe Streets already operates a regional
near-miss reporting form, modeled on what other cities — including
Charlottesville — run through their own transportation departments.
Today, through a formal partnership with DASH, QR codes linking to the
NVFSS form (and soon SMS short codes supporting WhatsApp as well) are
posted at Alexandria bus stops. The data is real, the data is
published, and the partnership works.
What's missing is reach. A single transit partnership, however
successful, isn't a citywide signage strategy. The same signage that works
on a DASH bus stop would work on a city-maintained crosswalk pole at a
high-incident intersection, on construction signage where a sidewalk has
been closed, at the entrance to a city park, at a recreation center, at a
public school's perimeter signage, at a Capital Bikeshare station.
The construction-signage case is especially direct. A pedestrian forced
into the street by a closed sidewalk and then hit by a driver — one
of the contributing factors at Sanger and Van Dorn this April — is
exactly the kind of incident a near-miss form could have surfaced earlier,
if anyone walking past that closed sidewalk had known how to raise it.
Signage like this could make it easy.
And expanding the signage reaches people that the other two data layers
miss entirely. Residents whose first language isn't English. Residents with
reasons to avoid law enforcement contact. Residents who simply don't know
that “reporting a near-miss” is something they're allowed to
do. Those are exactly the residents whose experience is currently absent
from other Vision Zero datasets. Reaching them isn't a nice-to-have; it's
the data integrity question at the center of an equity-focused safety
program.
One thing this is not: a substitute for the reportable crash data
Alexandria already tracks. Near-miss reports are a complement —
supplementary input for specific corridor projects, not a citywide
prioritization tool. The participation-bias concern with crowdsourced
reports is real: if reporting concentrates in Rosemont and underrepresents
the West End, the data should never outweigh hard crash data when
allocating resources. The expansion strategy above exists precisely to
broaden who participates, not to elevate near-miss reports above crash
records.
A resident asking “how many people were hit by drivers in
Alexandria last month?” should be able to get an answer within
days, not months. A resident reporting a close call on a
sidewalk that shouldn't have been closed should not have to know which
advocacy organization runs the form. A pedestrian struck and refusing
medical attention should still register somewhere as a person Alexandria
failed to keep safe.
That's the data infrastructure a Vision Zero strategy
depends on. It's also what residents have a right to expect.
A note on access
This analysis was only possible because Alexandria's primary dispatch
radio is broadcast on an open frequency and archived publicly on OpenMHz.
That isn't the case everywhere. A growing number of US localities have
moved to encrypted dispatch in recent years. The reasons cited are usually
officer-safety; the practical effect is that the public loses its only
upstream view of what is actually happening on city streets. If this
analysis prompts a similar conversation in Alexandria, that would be a
shame — and counterproductive. The audio is the evidence. Removing
it doesn't fix the gap between what dispatch hears and what the city
officially counts; it just removes the public's ability to see it.
I have worked hard to model good stewardship of this data. License plate
numbers, witness contact information, and other identifying details that
came over the air have been muted in the audio published here, and
redacted in the transcripts. The originals were broadcast publicly. The
choice to scrub them is mine, in the interest of demonstrating that public
dispatch access and individual privacy can coexist.
They can. This piece is intended as evidence that they do.
Alexandria isn't the first city where this has been studied
The DC findings were not evenly distributed. The highest rates of
unreported crashes were east of the Anacostia River — in Wards 7 and
8,
lower-income and majority-Black wards
— where the official record was least complete in exactly the places
that needed accurate data the most. The DC Policy Center concluded that
DDOT was making infrastructure decisions on data that systematically
undercounted pedestrian harm in the wards that already had the least
political voice over those decisions.
Their work continues. In January 2025, the DC Highway Safety Office
awarded a
$470,000 federal grant
to UVA Darden professor Chris Parker, Charlotte Jackson, and a research
team to extend the methodology — adding predictive modeling, expanded
data sources, and a dashboard that maps the streets students must cross to
get to school. Their findings now suggest that between 30 and 40 percent
of crashes go unreported in DC's official data, and the equity
disparity remains the central concern.
What this means for Alexandria: the data-infrastructure question this
piece raises is being actively studied right across the river, with
federal funding, by researchers whose methodology I independently
replicated when I built the tool behind this piece. Alexandria is
conspicuously absent from that regional conversation. The incidents
documented below are the first systematic application of this approach to
an Alexandria-sized Virginia city. There is no reason to believe
Alexandria's data is more complete than DC's — and good reason, given
the same statutory reporting threshold and the same dispatch infrastructure,
to expect the gap here is similar.
How this was built and what it can and can't show
This methodology is not novel. The approach — pulling
dispatch audio, transcribing it, and comparing flagged incidents against
official crash data — was developed by Emilia Calma and Charlotte
Jackson at the DC Policy Center in 2021, and is being extended in 2025
under federal grant funding by a research team at UVA Darden and the DC
Pedestrian Advisory Council. I built the tool behind this piece
independently before learning of the DC work; the convergence is itself a
signal that the underlying approach is sound. What follows describes how
this specific implementation was built and what it can and cannot show.
The tool. A Python program polls
OpenMHz
for new recordings from Alexandria Police and Fire dispatch talkgroups,
transcribes each with OpenAI's Whisper, and flags recordings whose transcripts
match a keyword list for pedestrian, cyclist, scooter, or crash activity on
Mount Vernon Avenue and Braddock Road. A human — me, for this review
— reviewed every flagged recording and marked it real, false, or
unclear. All 28,321 recordings from April 1–23 went through the same
pipeline; 129 flagged; 73 incident clusters produced; this piece features 18
Alexandria-city pedestrian/cyclist/scooter strikes, plus one additional
incident involving Alexandria units handling a Virginia State Trooper strike
on I-495 at Telegraph Road on April 7 (covered separately
by WTOP).
Scope. April 1 through April 22, 2026. Alexandria, Virginia.
Pedestrian, cyclist, and scooter strikes that generated a dispatch call on
primary police (APD 1) or fire/EMS (AFD 2 Alpha) talkgroups. Does not
include incidents on secondary tactical or incident-command channels.
Exclusions. Incidents that turned out on investigation not to be
strikes are excluded. One call during the review period came in reporting a
cyclist struck by a driver; the responding officer on scene clarified that
a passenger had fallen off the back of the bike — that call is not
among the incidents below. Five unrelated calls that the flag logic
triggered (a parked-vehicle strike, cars driving on a pedestrian bridge,
three car-on-car crashes) were excluded because they were not
pedestrian/cyclist/scooter incidents.
What this can't see. Follow-up investigation detail, witness
statements, final injury assessments, and whether charges were ultimately
filed do not flow through dispatch. Some incidents may on full investigation
turn out less serious than they sounded on the radio, and the reverse is
also true: dispatch often has only sparse initial information on serious
incidents. A handful of corrections in either direction does not close the
gap between 6 and 18.
Known omission. At least one Alexandria pedestrian incident in
this window did not go over primary dispatch radio at all — a
predawn incident on April 1 at Mill Road and Eisenhower Avenue that I
learned about independently through advocacy contacts. Some fraction of
pedestrian-at-fault or walk-in-reported incidents may be handled through
channels this tool cannot see. The 18 figure is a count of what the tool
caught and I verified; the true count of pedestrian incidents in Alexandria
during this window is almost certainly higher.
Where the data lives. The full corpus (28,321 transcripts, 129
flagged transmissions, 73 incident clusters) is preserved in the underlying
dataset. Raw audio for every incident below was downloaded during the
review window so it survives the 30-day OpenMHz archive rollover.
The tool itself is open source. If you're an advocate, journalist,
or civic technologist in another city and want to do the same thing locally,
the pipeline is documented and adaptable at
github.com/ElenaH77/dispatch-pedestrian-monitor.
The README, an ADAPTING.md adaptation guide, and an ETHICS.md are the
three docs you'd start with.