Town of East Greenbush · Flock Safety license plate readers
This does not need to be exact to be useful. The uncertainty around the figure below is wide, and the precise number should not be relied on. What it establishes is the order of magnitude, and that holds under every assumption tested here: tens of millions of vehicles have passed these readers.
The PDF is the version to attach to an email. An emailed attachment becomes part of the recipient's own record; a link does not.
Everything the Town publishes about this program counts searches. Nothing published counts reads. The cameras read every plate that passes, continuously, and no Town record states how many that comes to. The detection records that would show it are subject to a thirty-day retention setting. An estimate from published state traffic data is what remains.
I am not a statistician or a traffic engineer. I am a private citizen who used NYSDOT traffic counts and the locations named in the Town's own Resolution 40-2025. Every figure here is arithmetic anyone can repeat, and the inputs are shown so that anyone who thinks a number is wrong can point to the step they disagree with. A professional would do this better, and the limits are set out in full below.
Vehicle passages past four of the five readers, October 1 2021 through November 18 2026, a period of 1,875 days of which 2021 contributes 92 and 2026 contributes 322. The fifth reader is not counted because no state traffic station exists on its road.
Positions computed from mapped coordinates and plotted to scale. Circle area is proportional to recorded passages. This is a scale plot, not a street map; roads are named rather than drawn.
| # | Location, as named in Resolution 40-2025 | Direction | NYSDOT station | AADT | × 1,875 days |
|---|
Every AADT above is a number NYSDOT published. The station was selected and the direction stated in the Town's own resolution was applied. No traffic count here was calculated, adjusted, smoothed or interpolated by me. Anyone disputing the counts themselves is disputing how the New York State Department of Transportation measures, factors and publishes traffic data. Every station dashboard is linked at the end so each number can be read off the state's own page. What is mine is the rest: which station to read for each reader, the single-direction share, the 1,875 day multiplier, and carrying 2025 across 322 days of 2026. Those choices are set out above so they can be argued with, and the installation date is the one most able to move the total.
As NYSDOT publishes them, totals in both directions. The figures used above are the single direction each reader faces, roughly half of each.
| Reader | Station | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|---|
| 01 Sewer plant | 140038 | 16,845 | 16,481 | 18,779 | 19,275 | 16,403 |
| 02 Genet School | 140102 | 13,902 | 13,325 | 15,152 | 13,279 | 13,166 |
| 03 Hewitt's | 140125 | 17,918 | 17,175 | 16,574 | 17,012 | 16,867 |
| 04 Grandview | 140131 | 26,002 | 24,561 | 20,633 | 21,178 | 24,125 |
NYSDOT labels every published year Actual or Estimated:
| Station | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| 140038 | Estimated | Actual | Estimated | Estimated | Actual |
| 140102 | Actual | Estimated | Estimated | Actual | Estimated |
| 140125 | Estimated | Estimated | Actual | Estimated | Estimated |
| 140131 | Estimated | Actual | Estimated | Estimated | Actual |
The headline figure uses the 2025 column, where two of the four are Actual. The year-by-year cross-check draws on the whole grid, where seven of the twenty values are Actual. Neither method rests entirely on measured years. No 2026 count has been published for any of these stations, so the 2025 figure is carried across the 322 days of 2026.
Reader 05 is not counted at all. The four readers that were removed are not counted, and no operating period is estimated for them, because no installation or deactivation dates have been produced. The lower directional figure is used at Hewitt's. The lower of two calculation methods is used. Nothing here counts a vehicle passing the same reader twice.
That makes the figure conservative. It does not make it a guaranteed minimum and it should not be called one: an AADT estimate can overstate as well as understate, and leaving readers out does not prevent the remaining four from being overstated. Every discretionary choice here was made in the direction that lowers the number.
Every location was visited, photographed and checked against a handheld GPS fix on 16 November 2026, in one continuous trip between 14:13 and 15:07. Coordinates below are the mapped positions used for the distance measurements in this document.
| # | Named in Resolution 40-2025 | Coordinates | Closest photo fix |
|---|---|---|---|
| 01 | Eastbound Turnpike at Sewer Treatment Plant | 42.6267889, −73.7417097 | 23 ft |
| 02 | Northbound Genet School and Route 4 | 42.6005802, −73.7086536 | 12 ft |
| 03 | Route 4 at Hewitt's | 42.6306937, −73.7011293 | 99 ft |
| 04 | Southbound Troy Road at Grandview Drive | 42.6458133, −73.6980601 | 10 ft |
| 05 | Westbound 3rd Avenue Extension at Krey Boulevard | 42.6442146, −73.7057620 | 215 ft |
The distances are how far away the photograph was taken, not the accuracy of the fix. Reader 05 stands on Third Avenue Extension about 150 ft east of the end of Patroon Lane; Patroon Lane, not Krey Boulevard, is the nearest intersection to it. Reader 05 is excluded from every total here, because NYSDOT publishes no traffic count for that road.
Resolution 40-2025 gives a direction of travel for four of the five readers. Route 4 at Hewitt's has none, so the lower of that station's two directional counts is applied.
That reader stands on Troy Road at the I-90 interchange. It was visited and photographed on 16 November 2026 and observed facing southeast, along the side carrying traffic toward the interchange, which there is southbound. Applying the southbound count would bring the total to 65,850,000. That figure is not used: the resolution states no direction, and where a published record and a personal observation could differ, this follows the record.
Reader 02, Genet Elementary. 66 ft from the school property corner and 203 ft from the school vehicle entrance, reading northbound Route 4. Vehicles approaching the school from the south on Route 4 pass it, school buses included.
Reader 03, Columbia High School. The reader stands on Troy Road at the I-90 interchange. The school, at 962 Luther Road, lies to the southeast. That campus has several approaches, so this is proximity and nothing more.
Reader 04, Woodland Hill Montessori. Reading southbound Troy Road at Grandview Drive. Grandview Drive is a dead end, recorded by NYSDOT as "GRANDVIEW DR from US 4 to DEAD END," and is the only road access to the school. Every vehicle leaving it must enter Troy Road at this intersection.
Resolution 40-2025 names the five locations and states no basis for choosing any of them. No siting analysis, no field of view documentation and no record of any consideration of proximity to schools has been produced by the Town.
A state count point sits on Grandview Drive itself, station 145030, which NYSDOT describes as "GRANDVIEW DR from US 4 to DEAD END." The state classes it 7U Local against 3U Principal Arterial for Route 4, records an 85th percentile speed of 21.9 mph and zero buses, and its only field count was taken in April 2017, four years before the cameras.
Resolution 40-2025 places reader 04 on Troy Road watching southbound Route 4, and the Grandview intersection falls inside station 140131's segment. Using the dead-end station would have put this reader at roughly 221,000 instead of 23,561,250.
The Town's transparency portal reports 57 search sessions in its most recent thirty-day window, as captured on 19 November 2026. Other agencies' published audit files show 18,513 searches run with East Greenbush credentials between 31 October 2023 and 10 November 2026. The median one reached 5,522 camera networks; the largest reached 97,836 individual cameras. Those are lookups against this pool.
Applying one year's count across five years invites an objection, so it was also run the other way: each calendar year's own published count against the days falling in that year. That gives 64,999,769, a difference of 0.6%. The lower figure is the one used here. Both draw on the same NYSDOT data, so this is not an independent check on the counts: it shows only that the choice of method barely moves the result.
NYSDOT publishes a directional split only for the current year, so the 2025 share was held constant across every year: 46.8% eastbound at 140038, 46.5% northbound at 140102, 48.1% northbound at 140125, 52.1% southbound at 140131. Holding it constant is a simplification, and the shares are given so the cross-check can be reproduced.
Three of the four stations come out higher under the year-by-year method and Grandview comes out lower, so the differences largely cancel. Those shares are printed to one decimal place, so anyone recomputing from the published yearly counts will land within about twenty thousand passages of 64,999,769 rather than exactly on it, roughly three hundredths of one percent. At this scale that is a rounding artifact, not a disagreement.
NYSDOT has published no 2026 count for any of these stations. The 2025 figure is carried across the 322 days of 2026 in the period, which is 17.2% of the total. This is the largest single assumption here, and it is why the cross-check above matters.
No traffic station for reader 05. No locations on record for the four readers removed in 2022. No plans or maps, though Resolution 40-2025 refers to them. No Law Enforcement Equipment Agreement with NYSDOT, though the resolution directs that one be prepared.
Annual Average Daily Traffic is an estimate of the average vehicles crossing a point per day over a year. It is not a tally. NYSDOT derives it from short field counts, factored up for season and day of week, and interpolated for years with no count.
Each station opens directly on the NYSDOT Traffic Data Viewer, and each reader's coordinate opens on DeFlock. Both are linked from the table above. The rejected Grandview Drive station is linked here too, so the matching decision can be checked as readily as the figures.
| Reader 01 | 140038 |
| Reader 02 | 140102 |
| Reader 03 | 140125 |
| Reader 04 | 140131 |
| Rejected, Grandview Dr | 145030 |
Two of the four figures are the state's own estimates. The largest component is measured, and measured counts carry 58.7% of the result.
| 04 Grandview · 140131 · 36.5% | Actual |
| 03 Hewitt's · 140125 · 23.5% | Estimated |
| 01 Sewer plant · 140038 · 22.3% | Actual |
| 02 Genet · 140102 · 17.7% | Estimated |
58.7% of the total rests on measured counts. The largest single component, reader 04 at over a third of the whole figure, comes from an Actual count taken in February 2025 over 53 hours.
Reader 01 sits on Columbia Turnpike, which NYSDOT counts at several points. The station used reads 16,403. Its neighbours on the same road read 21,274 north and 14,502 southeast, so the one used sits between them.
| Neighbour north applied | 68,908,762 |
| Station used | 64,633,125 |
| Neighbour southeast applied | 62,964,477 |
The choice moves the total by about 9%, and the nearest alternative on the same roadway would have produced a larger figure, not a smaller one.
The counts informing the years used here are seven field counts across the four stations: August 2022 and June 2025 at 140038 (83 and 69 hours), September 2021 and July 2024 at 140102 (190 and 76), October 2023 at 140125 (72), and October 2022 and February 2025 at 140131 (86 and 53). 629 hours of measurement in all. Everything between them is interpolation.
The stations have longer histories, with counts going back to 2010 at two of them, but those predate the cameras and do not inform the years used here.
2021 contributes 92 days from October 1. 2026 contributes 322 days with no published count, so 2025 is carried forward. Together 22% of the period rests on a partial year and a carried figure. November 18 is the meeting date, so the last three days are projected, worth 103,413 passages or 0.16%.
This treats the five current readers as in place for the whole period, on the reading that the 2022 reduction to five removed four of the original nine rather than replacing them. The quantities are on the record: the order form of 27 September 2021 lists nine Flock Falcon Cameras, and an undated Amendment Order Form states that the Agency “will reduce Flock Falcon Cameras from 9 to 5” for a new initial fee of $13,750 and a recurring total of $12,500, the same figures Resolution 148-2022 approved on 15 June 2022. What is not on the record is when any of them was installed or taken down, so treating the five as in place throughout is an inference, and it is the assumption most able to move the total.
It is an inference because the first request of 16 July 2026, which the Town numbered 78 in its reply, asked for every device with its installation date, and the production of August 24 returned a camera list with no installation dates. The dates exist in Flock's activation records. Anyone wishing to correct the period can produce them.
This estimates vehicle passages. A camera can fail to read a plate through weather, speed, angle or obstruction, so the count of successful reads is a figure only Flock and the Town can produce. Repeat passages are counted each time, which is right for passages but overstates distinct vehicles.
| estimate reduced 10% | 58,169,812 |
| reduced 25% | 48,474,844 |
| reduced 50% | 32,316,562 |
| reduced 75% | 16,158,281 |
The estimate would have to be reduced by more than 84% to fall below ten million. Put another way, it is about six and a half times that threshold.
These are illustrative reductions, not a confidence interval. None is offered here, because establishing one would need NYSDOT's underlying estimation records rather than its published figures.
One record. The detection data reports actual reads per camera. It was requested on July 27 2026. Under the thirty-day retention setting the months in question would ordinarily have expired before the response period under the Freedom of Information Law had run. No record confirming deletion exists, only the retention setting the Town publishes.
Whether the Town itself could ever see that figure is unresolved, and no claim is made here that it could. The data sits on the vendor's systems, and Section 1.12 of Flock's current terms defines reports generated through the service as Flock Property with no right to export or retain copies. The Attorney to the Town was asked on September 4 2026 whether the agreement gives the Town a right to a complete audit record. No answer has come.