Before accessible pedestrian signals were installed, blind participants in a pre-post field study started crossing during the WALK interval on only 25% of crossings; after installation, that figure jumped to 84% (NCHRP Web-Only Document 117B, 2007). The stakes of getting the timing right are high: 7,314 pedestrians were killed on US roads in 2023, 18% of all traffic deaths (NHTSA, Pedestrians: 2023 Data). New York City alone is under court order to equip at least 10,000 signalized intersections with APS by the end of 2031, and it had reached 3,247 by the end of 2024 (NYC DOT, APS Program Status Report 2025). We aggregated data from NHTSA, the Governors Highway Safety Association, the US Access Board, the Transportation Research Board’s NCHRP program, NYC DOT, the City of Toronto, the UK Department for Transport, the World Health Organization, and the American Foundation for the Blind. For the reading side of non-visual access, see our companion roundup on braille literacy statistics.
TL;DR
- 7,314 US pedestrians were killed in 2023, 18% of all traffic deaths (NHTSA, Pedestrians: 2023 Data).
- US pedestrian deaths rose 49% in a decade, from 4,910 in 2014 to 7,314 in 2023 (NHTSA, Pedestrians: 2023 Data).
- 2025 pedestrian deaths are projected at 6,732, down 7% from 2024 (GHSA, Pedestrian Traffic Fatalities by State: 2025 Preliminary Data).
- At least 2.2 billion people worldwide have a near or distance vision impairment (WHO, Blindness and visual impairment fact sheet 2026).
- 50.18 million US adults reported trouble seeing or being blind in 2022 (AFB, citing NHIS 2022).
- With APS, crossings started during WALK rose from 25% to 84% (NCHRP Web-Only Document 117B, 2007).
- Mean starting delay for blind pedestrians fell from 5.1 to 2.9 seconds with APS (NCHRP Web-Only Document 117B, 2007).
- NYC had APS at 3,247 intersections as of December 31, 2024 (NYC DOT, APS Program Status Report 2025).
- NYC’s 2024 APS retrofits averaged about $60,800 per intersection (NYC DOT, APS Program Status Report 2025).
- Toronto had 1,219 signalized intersections with APS as of May 26, 2023 (City of Toronto, Accessible Pedestrian Signals page).
- PROWAG requires detectable warnings at least 24 inches deep with 0.2 inch high truncated domes (US Access Board, PROWAG Chapter 3, 2023).
- Hybrids were about 1.18 times more likely than combustion vehicles to hit a pedestrian, which drove the US quiet-car sound rule (NHTSA, FMVSS No. 141 Final Rule 2016).
1. The Pedestrian Safety Baseline: Where and When People Die
The national numbers explain why non-visual crossing cues are a safety tool and not a courtesy. Pedestrian deaths rose 49% in ten years, from 4,910 in 2014 to 7,314 in 2023, and the risk concentrates exactly where sight matters most: after dark and away from controlled crossings. A blind or low-vision traveler cannot see a driver’s headlights swing into a turn, so the information has to arrive by sound or touch.
The intersection share (17%) looks small next to the 74% of deaths at non-intersection locations, but that is partly the point: a signalized crossing with audible and vibrotactile cues is one of the few places where infrastructure can actively tell a pedestrian when it is their turn. Every crossing that gets pushed to an unmarked midblock spot because the signal is unusable is a crossing moved into the higher-risk column.
| Metric | Value | Source |
|---|---|---|
| Pedestrians killed in US traffic crashes, 2023 | 7,314 (18% of all traffic deaths) | NHTSA, Pedestrians: 2023 Data |
| Change vs 2022 | -3.7% (from 7,593) | NHTSA, Pedestrians: 2023 Data |
| Pedestrian deaths, 2014 vs 2023 | 4,910 vs 7,314 (+49%) | NHTSA, Pedestrians: 2023 Data |
| Share of pedestrian deaths in the dark | 77% (19% in daylight) | NHTSA, Pedestrians: 2023 Data |
| Share at non-intersection vs intersection locations | 74% vs 17% | NHTSA, Pedestrians: 2023 Data |
| Pedestrians injured, 2023 | 68,244 (+1.3% vs 2022) | NHTSA, Pedestrians: 2023 Data |
| Pedestrian fatality rate | 2.18 per 100,000 population | NHTSA, Pedestrians: 2023 Data |
| Frequency of pedestrian deaths | One every 72 minutes | NHTSA, Pedestrians: 2023 Data |
Context: NHTSA also reports that 84% of pedestrian deaths happened in urban areas, 70% of those killed were male, 26% of deaths occurred between 6 and 8:59 p.m., and alcohol (BAC of .01 g/dL or higher) was involved in 46% of fatal pedestrian crashes. Full tables are in the NHTSA Pedestrians: 2023 Data fact sheet. Most recent available final data: NHTSA, 2023.
2. The 2024-2025 Trend: Fewer Deaths, Same Risk Profile
The preliminary 2025 count is the first clearly good news in years, but the risk profile has not changed. GHSA projects 6,732 pedestrian deaths in 2025, down 7% from 2024 and the first fatality rate below 2.0 per 100,000 since before the pandemic, yet the total still sits 5% above the pre-pandemic baseline. Darkness, missing sidewalks and inattentive drivers remain the dominant patterns.
For blind and low-vision pedestrians, the distracted-driving trend matters most. A pedestrian who relies on hearing traffic to judge a gap cannot compensate for a driver who is not looking: the cue has to come from the infrastructure itself.
| Metric | Value | Source |
|---|---|---|
| Projected pedestrian deaths, 2025 | 6,732 | GHSA, Pedestrian Traffic Fatalities by State: 2025 Preliminary Data (July 2026) |
| Change vs 2024 | -7% | GHSA, 2025 Preliminary Data |
| Position vs pre-pandemic baseline | 5% above | GHSA, 2025 Preliminary Data |
| Share of 2024 pedestrian deaths after dark | 76.5% | GHSA, 2025 Preliminary Data |
| Share of 2024 deaths where no sidewalk was present | 62.2% | GHSA, 2025 Preliminary Data |
| Hit-and-run share of 2024 pedestrian deaths | 23.8% (1,669 deaths) | GHSA, 2025 Preliminary Data |
| Distracted driving share, 2024 vs 2019 | 21.7% vs 15.6% | GHSA, 2025 Preliminary Data |
| Share of 2024 deaths involving light trucks | 50.4% | GHSA, 2025 Preliminary Data |
Source: GHSA, Pedestrian Traffic Fatalities by State: 2025 Preliminary Data. GHSA figures are preliminary projections built from state highway safety office counts; NHTSA’s final FARS numbers for 2024 and 2025 may differ.
3. Who Relies on Audio and Tactile Cues
The user base for APS and tactile paving is far larger than the number of people who are totally blind. At least 2.2 billion people worldwide have a near or distance vision impairment, and for at least 1 billion of them it could have been prevented or has yet to be addressed (WHO). In the US, survey definitions swing the count enormously: a broad “trouble seeing” question in the National Health Interview Survey yields 50.18 million adults, while the narrower American Community Survey disability question yields about 7.7 million people.
Age is the multiplier. Visual disability prevalence was 6.6% among people 65 and older in the ACS 2016 data, three times the all-ages rate, and NHTSA counts 1,533 pedestrians among the 7,891 people aged 65+ killed in 2023 traffic crashes (19%). Aging populations mean more pedestrians whose vision is fading but who still walk to transit, shops and clinics. Many of the same people depend on refractive correction long before they need a cane, a shift tracked in our pediatric myopia progression statistics.
| Metric | Value | Source |
|---|---|---|
| People with near or distance vision impairment, worldwide | At least 2.2 billion | WHO, Blindness and visual impairment fact sheet (Feb 2026) |
| Cases that could have been prevented or are unaddressed | At least 1 billion | WHO, fact sheet 2026 |
| Annual global productivity cost of vision impairment | US$411 billion | WHO, fact sheet 2026 |
| US adults reporting trouble seeing or blindness | 50.18 million | AFB, citing NHIS 2022 |
| US people with a visual disability (ACS definition) | 7,675,600 (2.4%) | NFB, citing ACS 2016 via Cornell (Erickson et al., 2017) |
| Visual disability prevalence, age 65+ | 3,171,100 (6.6%) | NFB, citing ACS 2016 |
| Working-age adults with visual disability employed full-time, full-year | 29.5% | NFB, citing ACS 2016 |
| Working-age adults with visual disability below poverty line | 27.7% | NFB, citing ACS 2016 |
Sources: WHO fact sheet, AFB statistics and NFB blindness statistics. Most recent available ACS figures on the NFB page: 2016. Low employment matters here because people who cannot drive depend on walking and transit to reach work.
4. What APS Actually Change at the Curb
The research case for audible signals is unusually concrete. Without them, blind pedestrians time crossings by listening for the surge of parallel traffic, a cue that fails at quiet intersections, with right-turn-on-red traffic or with leading pedestrian intervals. In a pre-post study summarized by NCHRP, blind participants started crossing during WALK on only 25% of crossings before APS installation and on 84% afterwards, and crossings that finished after perpendicular traffic got its green fell from 50% to 12%.
APS also fix a wayfinding problem, not just a timing one. Pushbutton locator tones pull travelers to the right spot: starting within the crosswalk rose from 70% to 97% with APS in SKERI research. That is the difference between crossing inside the marked lane and drifting into turning traffic.
| Metric | Value | Source |
|---|---|---|
| Crossings started during WALK, before vs after APS | 25% vs 84% | NCHRP Web-Only Document 117B (2007), NEI Portland pre-post study |
| Crossings completed after perpendicular green, before vs after APS | 50% vs 12% | NCHRP Web-Only Document 117B (2007) |
| Mean starting delay, without vs with APS | 5.1 s vs 2.9 s | NCHRP Web-Only Document 117B (2007) |
| Crossings started within the crosswalk, without vs with APS | 70% vs 97% | NCHRP Web-Only Document 117B (2007), SKERI research |
| Crossings begun from outside the crosswalk by travelers who located it independently | About 29% | NCHRP Web-Only Document 117B (2007) |
| Trials where blind pedestrians asked for help knowing when to start (no APS) | 24% | NCHRP Web-Only Document 117B (2007), SKERI research |
| Study participants in the NCHRP 3-62 field tests | 40 (20 in Tucson, 20 in Charlotte) | NCHRP Web-Only Document 117B (2007) |
Source: NCHRP Web-Only Document 117B, Guidelines for Accessible Pedestrian Signals: Final Report. Most recent available data: TRB/NCHRP, 2007. These are still the reference field studies behind US APS guidance, but they predate today’s signal timing practices and quieter vehicle fleets; new controlled studies are overdue.
5. APS Rollout and Cost: The New York City Benchmark
New York is the most closely tracked APS program in North America because a federal court sets its annual targets. NYC DOT has beaten every court benchmark so far, equipping 2,284 intersections in 2022-2024 against a required 1,600, and reached 3,247 APS intersections citywide by December 31, 2024. The harder part is still ahead: the schedule climbs to about 1,100 intersections a year from 2027 to 2031.
Cost is the real constraint. A single APS unit on an existing pole has historically cost about $1,800, but 2024 retrofits averaged about $60,800 per intersection because many corners need new poles, wiring and controller work. Toronto, which upgrades 20 to 30 existing intersections a year, shows the pace a city reaches without a court order.
| Metric | Value | Source |
|---|---|---|
| NYC intersections with APS, as of Dec 31, 2024 | 3,247 | NYC DOT, APS Program Status Report (Dec 2025) |
| NYC intersections equipped, 2022 / 2023 / 2024 | 520 / 888 / 876 (required: 400 / 500 / 700) | NYC DOT, APS Program Status Report |
| Court target | At least 10,000 by end of 2031; all remaining signalized intersections by 2036 | NYC DOT, APS Program Status Report |
| Required annual pace, 2025 and 2026 | 900 intersections per year | NYC DOT, APS Program Status Report |
| Baseline cost per APS unit on an existing pole | About $1,800 | NYC DOT, APS Program Status Report |
| Minimum estimated cost per intersection | At least $14,400 | NYC DOT, APS Program Status Report |
| 2024 retrofit cost per intersection, average vs median | About $60,800 vs $52,000 | NYC DOT, APS Program Status Report |
| Toronto signalized intersections with APS, May 26, 2023 | 1,219 (20 to 30 existing intersections upgraded per year) | City of Toronto, Accessible Pedestrian Signals page |
Sources: NYC DOT APS Program Status Report and City of Toronto, Accessible Pedestrian Signals. NYC DOT puts spending over the three years of its installation contract at about $147 million. Toronto’s design is also a reminder that sound design varies by country: it uses a cuckoo sound for north-south crossings and a chirp or melody for east-west, and users hold the pushbutton for at least three seconds to activate the audible signal at most intersections.
6. Tactile Paving and APS Technical Standards
Standards turn good intentions into measurable surfaces and sounds. The US Access Board published its final Public Right-of-Way Accessibility Guidelines (PROWAG) on August 8, 2023, effective September 7, 2023, after a rulemaking that drew roughly 600 comments from 460 commenters on the 2011 proposal. PROWAG fixes truncated domes at 0.2 inches high with 1.6 to 2.4 inches between centers, and detectable warnings must run at least 24 inches deep.
The audio side is just as specific: a locator tone no longer than 0.15 seconds, repeating once per second, no more than 5 dBA above ambient, and a percussive walk tone at eight to ten ticks per second with a dominant 880 Hz component. The UK takes a different approach to tactile vocabulary, with six distinct surface types and red blister paving reserved for controlled crossings only.
| Metric | Value | Source |
|---|---|---|
| Truncated dome base diameter | 0.9 to 1.4 inches (23 to 36 mm) | US Access Board, PROWAG R305.1.1 (2023) |
| Truncated dome height and spacing | 0.2 inches high; 1.6 to 2.4 inches center to center | US Access Board, PROWAG R305.1.1 and R305.1.2 |
| Minimum detectable warning depth | 24 inches (610 mm) | US Access Board, PROWAG R305.1.4 |
| APS locator tone | 0.15 s or less, repeating at 1 s intervals | US Access Board, PROWAG R307.8.1 |
| APS locator tone volume | Up to 5 dBA above ambient; automatic adjustment capped at 100 dBA | US Access Board, PROWAG R307.8.2 |
| APS audible walk indication | 8 to 10 ticks per second, dominant component at 880 Hz | US Access Board, PROWAG R308.3.1 |
| UK tactile surface types | 6 | UK DfT, Guidance on the use of tactile paving surfaces (2022) |
| UK blister paving depth at controlled vs uncontrolled crossings | 1,200 mm vs 800 mm (blisters 5 mm high) | UK DfT, Guidance on the use of tactile paving surfaces (2022) |
Sources: PROWAG Chapter 3, Technical Requirements, the PROWAG overview and UK DfT, Inclusive mobility: using tactile paving surfaces. The Access Board lists adoption of PROWAG by GSA on July 3, 2024 and by the Department of Transportation on December 18, 2024. The UK corduroy hazard warning surface uses 6 mm high bars, 20 mm wide at 50 mm centers, typically 800 mm deep. On the US traffic-control side, FHWA published the MUTCD 11th Edition in December 2023, now current as Revision 1 (December 2025), per FHWA’s MUTCD site.
7. Vehicle Sound and Digital Cues: The Other Half of the Audio Picture
Crossing signals only cover the moment a pedestrian steps off the curb; the vehicles themselves are the other audio cue. NHTSA found hybrids were approximately 1.18 times more likely than combustion vehicles to be involved in a collision with a pedestrian, and the Pedestrian Safety Enhancement Act of 2010 ordered a minimum-sound standard. The result, FMVSS No. 141, requires an alert sound at speeds up to 30 km/h and reached full compliance on September 1, 2019.
The cost-benefit math was lopsided: under $75 per vehicle against an estimated 2,390 fewer injured pedestrians and pedalcyclists per year. Research is now pairing APS with digital localization: a 2022 study placing Bluetooth beacons at an intersection located users to the right crossing zone with 97.7% accuracy across nine zones. Synthetic speech is part of the same toolkit, from talking walk messages to navigation apps; our speech synthesis accessibility statistics cover that side.
| Metric | Value | Source |
|---|---|---|
| Hybrid vs combustion vehicle pedestrian-crash likelihood | About 1.18 times | NHTSA, FMVSS No. 141 Final Rule (2016) |
| Speed range covered by alert-sound requirement | Up to 30 km/h | NHTSA, FMVSS No. 141 Final Rule (2016) |
| Compliance dates | 50% of production by Sept 1, 2018; full by Sept 1, 2019 | NHTSA, FMVSS No. 141 Final Rule (2016) |
| Estimated injuries avoided (MY2020 analysis) | About 2,390 fewer injured pedestrians and pedalcyclists | NHTSA, FMVSS No. 141 Final Rule (2016) |
| Estimated cost per vehicle | $74.36 (passenger car), $71.97 (light truck) | NHTSA, FMVSS No. 141 Final Rule (2016) |
| Monetized benefits (MY2020) | $320.0 million (3% discount) / $247.5 million (7% discount) | NHTSA, FMVSS No. 141 Final Rule (2016) |
| BLE beacon crossing-zone localization accuracy | 99.8% (4 zones), 97.7% (9 zones) | Shin et al., Sensors (2022) |
Sources: NHTSA FMVSS No. 141 final rule (Federal Register, Dec 14, 2016) and Shin et al., Outdoor Localization Using BLE RSSI and Accessible Pedestrian Signals, Sensors 2022. Most recent available regulatory analysis: NHTSA, 2016; benefit estimates assumed electric and hybrid vehicles at 4% of registrations. For deaf-blind travelers, vibrotactile pushbuttons complement relay and text access covered in our telecommunications relay service statistics.
Summary: Tactile Paving and APS Audio by the Numbers
| Metric | Value | Source |
|---|---|---|
| US pedestrian deaths, 2023 | 7,314 (18% of traffic deaths) | NHTSA, Pedestrians: 2023 Data |
| Decade change in pedestrian deaths, 2014-2023 | +49% | NHTSA, Pedestrians: 2023 Data |
| Pedestrian deaths in the dark, 2023 | 77% | NHTSA, Pedestrians: 2023 Data |
| Pedestrians aged 65+ among 65+ traffic deaths, 2023 | 19% (1,533 of 7,891) | NHTSA, Pedestrians: 2023 Data |
| Projected pedestrian deaths, 2025 | 6,732 (-7%) | GHSA, 2025 Preliminary Data |
| 2024 deaths where no sidewalk was present | 62.2% | GHSA, 2025 Preliminary Data |
| People with vision impairment worldwide | At least 2.2 billion | WHO, fact sheet 2026 |
| US adults reporting trouble seeing or blindness | 50.18 million | AFB, citing NHIS 2022 |
| Crossings started during WALK, before vs after APS | 25% vs 84% | NCHRP Web-Only Document 117B (2007) |
| Mean starting delay, without vs with APS | 5.1 s vs 2.9 s | NCHRP Web-Only Document 117B (2007) |
| Crossings started within the crosswalk with APS | 97% (from 70%) | NCHRP Web-Only Document 117B (2007) |
| NYC APS intersections, end of 2024 | 3,247 | NYC DOT, APS Program Status Report |
| NYC court target | 10,000 intersections by 2031 | NYC DOT, APS Program Status Report |
| NYC average 2024 retrofit cost per intersection | About $60,800 | NYC DOT, APS Program Status Report |
| Toronto APS intersections, May 2023 | 1,219 | City of Toronto |
| PROWAG minimum detectable warning depth | 24 inches | US Access Board, PROWAG R305.1.4 |
| PROWAG locator tone volume cap | 5 dBA above ambient | US Access Board, PROWAG R307.8.2 |
| UK tactile surface types | 6 | UK DfT, tactile paving guidance 2022 |
| Hybrid pedestrian-crash likelihood vs combustion vehicles | About 1.18 times | NHTSA, FMVSS No. 141 Final Rule (2016) |
| Estimated cost of quiet-car alert per passenger car | $74.36 | NHTSA, FMVSS No. 141 Final Rule (2016) |
Methodology and Sources
- NHTSA, Pedestrians: 2023 Data (DOT HS 813 727, June 2025): crashstats.nhtsa.dot.gov
- GHSA, Pedestrian Traffic Fatalities by State: 2025 Preliminary Data (July 2026): ghsa.org
- WHO, Blindness and visual impairment fact sheet (February 2026): who.int
- American Foundation for the Blind, statistics (NHIS 2022): afb.org
- National Federation of the Blind, blindness statistics (ACS 2016 via Cornell, Erickson et al. 2017): nfb.org
- TRB, NCHRP Web-Only Document 117B, Guidelines for Accessible Pedestrian Signals: Final Report (2007): onlinepubs.trb.org
- NYC DOT, Accessible Pedestrian Signals Program Status Report (December 2025, data through 2024): nyc.gov
- City of Toronto, Accessible Pedestrian Signals: toronto.ca
- US Access Board, PROWAG overview and Chapter 3 Technical Requirements (final rule 2023): access-board.gov/prowag and technical chapter
- US Access Board, PROWAG final rule, Federal Register 2023-16149 (comment counts): govinfo.gov
- UK Department for Transport, Guidance on the use of tactile paving surfaces (January 2022): gov.uk
- FHWA, MUTCD 11th Edition: mutcd.fhwa.dot.gov
- NHTSA, FMVSS No. 141 Minimum Sound Requirements for Hybrid and Electric Vehicles, final rule (December 2016): govinfo.gov
- Shin K. et al., Outdoor Localization Using BLE RSSI and Accessible Pedestrian Signals for the Visually Impaired at Intersections, Sensors (2022): doi.org
- Data watch: There is no public national count of US intersections with APS or of curb ramps with compliant detectable warnings, so city programs (NYC, Toronto) stand in as benchmarks. US vision-loss counts vary by an order of magnitude with the survey question (NHIS trouble seeing: 50.18 million adults; ACS visual disability: about 7.7 million people), so the two figures are not comparable. The APS effectiveness studies (NCHRP, 2007) and the quiet-vehicle cost-benefit analysis (NHTSA, 2016) are older than three years and remain the most recent available. GHSA 2025 figures are preliminary projections and NHTSA final FARS data for 2024 will supersede the 2024 risk-factor shares. NFB’s ACS figures date from 2016. The NHTSA and TRB PDFs were read through a plain-text mirror and are cited at their original URLs.
Last updated: October 4, 2026. We update this roundup quarterly, and the next refresh is expected when NHTSA publishes Pedestrians: 2024 Data and NYC DOT releases its next APS Program Status Report.