The research, published in Frontiers in Future Transportation, was conducted by a team spanning the University of Palermo, Monash University Accident Research Centre in Australia, and Shenzhen University in China. Led by Muhammad Yasir of the University of Palermo, the study examined 141 pedestrian crashes recorded by the Palermo Municipal Police between 2015 and 2022, comprising 54 fatal crashes and 87 serious injury crashes. Rather than relying on posted speed limits or average travel speeds alone, the team used percentile travel speeds — the 15th, 50th, and 85th percentiles — drawn from TomTom Move, a traffic analytics platform built on anonymized GPS data from vehicles across the road network.

The choice of percentile speeds is technically significant. Impact speed, the velocity at which a vehicle actually strikes a pedestrian, is the most direct determinant of injury severity, but real-world crash data rarely capture it accurately. Police reports tend to be biased toward severe outcomes, and individual crash-involved vehicles may have been traveling far above or below the prevailing traffic speed. Percentile speeds offer a workaround: they characterize the full distribution of driver behavior at a crash location, capturing the slower, typical, and faster ranges of the speed environment. Prior research has shown that mean speed alone is a poor predictor of traffic injuries and that crash severity varies significantly with the speed distribution.

To make the speed data meaningful, the researchers extracted only the one-hour interval corresponding to each recorded crash time from a month of TomTom observations covering all hours and days of the week. If a crash occurred at 12:45 pm, only the 12:00 to 1:00 pm speed data was used. For the 54 fatal crash locations, estimates were derived from 29,208 route observations; for the 87 serious injury locations, from 41,938 observations. The team then applied established ordered logit risk functions, using the parameter estimates for the 15-to-59 age group as a broad reference category, to estimate the probability of fatal and serious injury (FSI) at each percentile speed.

The results reveal a steep, positive relationship between speed and pedestrian injury risk. At fatal crash locations, the estimated fatality risk was 0.7 percent at the 15th percentile speed (mean 20.1 km/h), 2.6 percent at the 50th percentile (mean 30.5 km/h), and 8.2 percent at the 85th percentile (mean 41.5 km/h). For serious injury crashes, the risks were far higher: 13.0 percent, 26.8 percent, and 43.6 percent at the 15th, 50th, and 85th percentile speeds respectively. Notably, the gap between the 50th and 85th percentiles was much larger than between the 15th and 50th, suggesting that the faster end of the driver speed distribution — where drivers may struggle to adjust to sudden changes in traffic conditions — drives a disproportionate share of severe pedestrian outcomes.

The counterfactual analysis is where the findings become actionable. The researchers mathematically reduced the entire percentile speed distribution at each crash location by 3 km/h and then by 5 km/h, and recalculated FSI risk. A 3 km/h reduction cut average fatality risk by 30.5 percent at the 15th percentile, 26.1 percent at the 50th, and 21.1 percent at the 85th. For serious injuries, the reductions were 23.0, 17.5, and 11.6 percent respectively. With a 5 km/h reduction, the benefits grew dramatically: fatality risk fell by 45.6 percent at the 15th percentile, 40.0 percent at the 50th, and 32.6 percent at the 85th, while serious injury risk dropped by 36.0, 28.1, and 19.4 percent.

These numbers carry particular weight in the Italian context. In 2024, Italy recorded 470 pedestrian fatalities, the second-highest in the European Union after Romania, and 72 percent of Italian pedestrian deaths involved cars as the main collision opponent — exceeding the EU average of 65 percent. According to the Italian National Institute of Statistics, pedestrians experience a fatality rate of 2.7 per 100 crashes, 4.4 times higher than car occupants. Palermo itself recorded 1,999 pedestrian crashes between 2015 and 2022, and its historic urban layout, dense pedestrian activity, and heavy reliance on private motorized travel — cars and motorcycles account for 69.2 percent of weekday trips — make it a textbook environment for vehicle-pedestrian conflict.

The study also documents the consequences of non-compliance. At one fatal crash site on Viale Ercole, where the posted limit is 50 km/h, TomTom data recorded 15th, 50th, and 85th percentile speeds of 52, 70, and 92 km/h during a one-month sample — evidence of frequent and substantial speed limit violations. The authors argue that intelligent enforcement tools such as speed cameras could help ensure compliance with reduced limits, and they point to the Movement and Place framework, which classifies streets by their mobility and social functions, as a way to justify lower limits in pedestrian-priority zones such as historic centers, school districts, and tourist areas. Palermo has already introduced limited traffic zones, pedestrian areas, and 30 km/h zones, and has pedestrianized more than 340,000 square meters of its historic center.

The findings align with a growing international evidence base. A Toronto study found that reducing speed limits from 40 to 30 km/h resulted in a 28 percent reduction in pedestrian motor vehicle collisions, while a Seattle case study showed that lowering limits by roughly 8 km/h led to a 17.2 percent decrease in crash severity and a 19.9 percent reduction in crash frequency. Traffic calming measures in Ghana, pedestrian hybrid beacons in the United States, and enhanced crosswalk lighting have all demonstrated complementary pathways to slower, safer streets. The Palermo study adds a crucial technical dimension: because it uses percentile speeds rather than single-dimensional measures, it captures the effect of speed variability across the distribution — showing that reductions applied across the entire speed environment, not just at the fastest end, yield meaningful safety gains.

The authors acknowledge limitations. Pedestrian ages were recorded only as a broad binary category, preventing age-specific risk estimation, and the injury risk model was not calibrated with Palermo or Italian crash data, so the estimated risks should be read as comparative estimates within the counterfactual framework rather than locally calibrated absolute probabilities. The TomTom speeds, derived from floating car data, were not independently validated against on-site radar measurements, though prior Italian research comparing commercial floating car data with fixed radar found reasonable estimates of urban travel speeds. Still, the core message is unambiguous: in dense historic cities where pedestrians and cars share constrained streets, shaving a few kilometers per hour off the entire traffic speed distribution could prevent a substantial share of the deadliest and most life-altering crashes. For urban planners and policymakers, the study suggests that the cheapest life-saving intervention may not be new infrastructure at all, but simply slowing down what already moves through the city.

Subject of Research: A study of vehicle-pedestrian crashes in Palermo, Italy, quantifies how percentile travel speeds affect pedestrian injury risk and shows that modest speed reductions of 3 to 5 km/h could meaningfully reduce fatal and serious pedestrian injuries.

Article Title: Evaluating the role of speed reduction on the severity of vehicle-pedestrian crashes: a case study in Palermo, Italy

Article References: Yasir, M., Logan, D., Migliore, M., D’Orso, G., Pervez, A., & Koppel, S. (2026). Evaluating the role of speed reduction on the severity of vehicle-pedestrian crashes: a case study in Palermo, Italy. Frontiers in Future Transportation, 7, Article 1805850. https://doi.org/10.3389/ffutr.2026.1805850

Image Credits: AI Generated

DOI: 10.3389/ffutr.2026.1805850

Keywords: pedestrian safety, vehicle speed, speed reduction, traffic crashes, Palermo, Italy, injury severity, road safety, percentile speeds, traffic calming, urban mobility, counterfactual analysis