
Road Safety
How AI and Smart Tech Are Reimagining Traffic Safety Across U.S. Cities
Actionable takeaways for city planners and policy makers aiming to create safer streets
Road safety is a top priority for U.S. cities in 2025. With traffic fatalities still alarmingly high, many cities are embracing a mix of cutting-edge AI technologies and proven street design measures to protect pedestrians, cyclists, and drivers. This educational overview will highlight how urban areas are deploying smart traffic systems and data analytics alongside traditional solutions like raised crosswalks, speed cameras, and roundabouts. We’ll look at real-world city examples, explain how each technology improves safety, and see how data-driven platforms help local governments make evidence-based decisions.
The goal is an accessible, engaging look at what’s working on America’s roads – and actionable takeaways for city planners and policy makers aiming for safer streets.
AI-Powered Solutions for Safer Streets
Modern cities are increasingly turning to artificial intelligence (AI) and smart infrastructure to prevent crashes before they happen. These technologies range from adaptive traffic lights that “think” on the fly to algorithms that flag dangerous hotspots. Here are some of the AI-based road safety innovations rolling out in 2025:
Smart Traffic Signals: Intelligent Intersections Reduce Crashes
Imagine a traffic light that adapts to real-time conditions – that’s the idea behind smart traffic signals. These AI-powered systems use sensors and cameras to detect cars, cyclists, and pedestrians and adjust light timing dynamically. Instead of fixed schedules, signals turn green or red based on actual demand, helping prevent the backups and risky behaviors that lead to accidents. For example, if pedestrians are waiting or a cyclist is detected in a bike lane, the smart signal can extend the walk phase or give a protected turn, reducing conflict. And if an ambulance is approaching, connected signals can turn green to clear a path.
Real-world example: Pittsburgh’s smart signal network, called SURTRAC, was one of the first. It uses AI developed at Carnegie Mellon University to optimize 50+ intersections. The results were impressive – travel times dropped by 25% and idling by 40%, and officials note that better timing has curbed red-light running. Los Angeles has also upgraded over 4,500 traffic lights with adaptive timing, saving an estimated 9.5 million hours of driver delay each year. While these systems often target congestion and emissions, they improve safety too: smoother flow means fewer rear-end crashes, and giving pedestrians a head start or detection can prevent collisions. Even smaller cities are testing AI at intersections. In 2024, Seattle partnered with Google on a pilot at three intersections – the AI analyzes smartphone navigation data to adjust signals in real time. The goals are to reduce stop-and-go waves (which can cause sudden braking crashes) and cut idling, all while improving safety by coordinating green lights. Cities like Baltimore and Nashville are not far behind – Nashville plans a $158 million initiative to modernize nearly 600 signals with cameras and AI by the 2030s, explicitly aiming to both speed up traffic and reduce accidents with the latest tech.
Predictive Analytics & Geospatial AI: Anticipating Crashes Before They Occur
One of the most promising uses of AI is crunching data to predict where crashes are likely and guide proactive fixes. Instead of waiting for a tragic accident to occur, city agencies now analyze patterns – speed data, past collisions, near-miss reports – to forecast trouble spots. These predictive crash analytics use machine learning and geospatial AI to flag “hot zones” so officials can intervene (through engineering or enforcement) before the next crash.
Real-world example: Hawaii’s Department of Transportation (HDOT) built a safety analytics platform using AI to take this proactive approach. The system ingests statewide crash data and driving patterns and then maps out predicted collision numbers for each location. Rather than just reacting to past crashes, HDOT’s tool forecasts where 18,100 crashes could happen over a three-year period and even estimates how many lives could be saved with the right countermeasures. Crucially, it doesn’t stop at prediction – it also recommends solutions. For example, the AI might notice a pattern of high-speed incidents near a crossing and suggest installing a raised crosswalk or better lighting there. Hawaii’s platform has a “Vision Zero” mindset: it shifted the focus to positive metrics like lives saved and speeds reduced instead of just counting past injuries. Early results are encouraging – one forecast estimated 6.9 lives saved over three years by applying the AI’s recommendations (things like rumble strips, new signage, and raised pedestrian crossings in key areas). HDOT now uses these analytics maps to secure funding and explain to the public why certain safety projects are priorities.
At Urban SDK, we provide an integrated geospatial AI platform that helps cities analyze traffic volumes, monitor real-time speeds, and map crash locations with pinpoint accuracy. This empowers transportation planners to prioritize high-risk areas and allocate resources where they’ll have the greatest safety impact. In Camas, Washington, we partnered directly with city officials to tackle speeding and collision hotspots. By aggregating real-time speed data and crash reports through Urban SDK, the city gained clear visibility into problem zones — enabling smarter enforcement strategies and roadway redesigns. Our platform allowed them to track whether new interventions like speed humps or police patrols were actually improving safety, using before-and-after analytics.
Connected Vehicle Infrastructure: Cars and Cities “Talk” for Safety
Another high-tech strategy gaining momentum is connected vehicle (V2X) technology, which enables cars, road sensors, traffic signals, and even pedestrians’ phones to communicate in real time. Imagine your car receiving a warning that a vehicle up ahead suddenly hit the brakes – or a traffic light “knowing” a speeding car is about to run the red and holding the crosswalk red a few extra seconds. Connected Vehicle Infrastructure makes these scenarios possible by creating a wireless dialogue between road users and infrastructure. The U.S. Department of Transportation sees huge safety potential here and in 2024 announced plans to accelerate V2X deployment nationwide. The goal: equip 20% of the National Highway System with V2X by 2028 (and 25% of traffic signals in the top 75 cities), on the way to full national coverage by 2036.
Real-world example: New York City led one of the biggest pilot programs testing connected tech. Over 3,000 city fleet vehicles (buses, trucks, cars) were fitted with devices to communicate with smart traffic signals in Manhattan and Brooklyn. When a driver was about to run a red light, the system sent a dashboard alert – and during the pilot this cut red-light violations by 41% at equipped intersections. Alerts for sharp curves and blind merges reduced dangerous speeding by an average of 8.7 mph on those stretches. The city also gave pedestrians a smartphone app that talks to the traffic signals. It can extend crossing time if someone hasn’t finished walking, or warn turning vehicles that a person is still in the crosswalk. In surveys, 83% of pedestrians said they felt safer using this connected tech while crossing the street. Following these successes, New York is expanding connected infrastructure as part of its Vision Zero initiative.
Other cities are exploring similar V2X safety benefits. In Phoenix, emergency responders have begun using connected vehicle systems to automatically trigger green lights as they approach intersections – shaving off response times and reducing the risk of collisions in emergency runs. Several state DOTs (in Arizona, Texas, Utah, and others) received federal grants in 2024 to install connected vehicle roadside units on busy corridors. These allow highway sensors to send drivers real-time alerts about hazards ahead (like sudden slowdowns, crashes, or even a pedestrian on a roadway). The technology can even detect a car that’s about to blow through a red light and warn other nearby drivers or cyclists to hit the brakes. While fully self-driving cars get a lot of buzz, it’s this connected tech – vehicles talking to each other and the city – that in the nearer term can prevent crashes. By 2025, we’re seeing more and more “smart corridors” and intersection pilots where infrastructure and vehicles share data to make streets safer for everyone.
Time-Tested Street Designs Saving Lives
Advanced technology alone won’t solve our road safety crisis – equally important are the street design and enforcement measures that we know work. Cities pioneering “Vision Zero” (eliminating traffic deaths) have found success with old-fashioned engineering: slowing cars down and designing conflicts out of the roadway. Here we highlight a few proven safety measures U.S. cities are rolling out or expanding in 2025:
Raised Crosswalks: Slowing Down for Pedestrians
A raised crosswalk is essentially a pedestrian crossing on a gentle speed hump. The crosswalk is elevated 3–6 inches above the road, forcing approaching drivers to slow to a safe speed – typically around 15–20 mph – as they go over the rise. This simple design tweak greatly increases pedestrian visibility (because drivers notice the ramp and the people on it) and reduces the severity of any collision that might still occur (lower speed = lower impact force). Raised crosswalks often double as traffic calming in busy pedestrian areas like school zones, downtown shopping streets, and trail crossings.
Real-world example: Honolulu, Hawaii is one city investing in raised crosswalks after identifying their benefits. Using the aforementioned AI analytics, Hawaii’s DOT pinpointed intersections where lowering driver speed would save lives, and has installed a series of raised pedestrian crosswalks at those locations. These are marked with bright paint and signs, and data shows they’re succeeding in getting drivers to hit the brakes. On the mainland, cities from Littleton, CO to Miami, FL have added raised crosswalks near transit stations and beach fronts to protect streams of foot traffic. While each location is unique, the general safety logic is universal: by creating a physical reminder for drivers to slow down (literally a bump in the road), raised crossings make it far more likely that a turning or approaching car can stop in time if someone is in the crosswalk. They also improve sightlines – as pedestrians stand a few inches higher and are more prominent. These are low-tech, relatively low-cost interventions, but when strategically placed they can virtually eliminate the kind of high-speed crosswalk crashes that too often prove fatal.
Speed Cameras: Automated Enforcement That Changes Driver Behavior
Speeding is a leading cause of severe crashes, and automated speed cameras have emerged as one of the most effective tools to rein it in. Unlike a police officer who can only catch so many violators, a speed camera system watches an area 24/7 and tickets any car exceeding the limit by a set amount. The mere presence of cameras has a strong deterrent effect – drivers know they will get caught and tend to slow down, even if they don’t see a police cruiser. Cities are increasingly deploying these cameras in school zones, high-crash corridors, and anywhere data shows speeding is a problem. The result: calmer traffic and lives saved.
Real-world example: New York City operates the largest speed camera program in the U.S., and it reported dramatic outcomes in a January 2025 report. At the over 2,000 camera locations, speeding violations dropped by 94% on average after cameras were installed. In other words, almost all drivers eventually learned to stick to the limit on those streets – a huge behavior shift. Because slower speeds mean fewer and less severe crashes, those areas also saw 14% fewer injuries and deaths compared to similar streets without cameras. New York found even extending camera operation to nights and weekends (previously they were off during those times) led to an additional 8% reduction in injuries during those hours. And importantly, most drivers only got one ticket and changed their habits, showing the enforcement is teaching lessons rather than repeatedly punishing the same people (though a small group of repeat offenders remains an issue).
Other cities have noted similar life-saving gains. On Philadelphia’s infamous Roosevelt Boulevard – one of the nation’s most dangerous urban roads – the introduction of speed cameras in 2020 cut fatalities in half within a few years. The city reported that excessive speeding dropped 90% on that arterial after camera enforcement began. Chicago and Washington, D.C. have likewise used speed cameras for over a decade and credit them with steady declines in crash rates, especially in school zones. The takeaway is clear: automated speed enforcement works. It slows drivers down even when police can’t be present, protecting everyone on the road. In 2025, many U.S. cities are expanding these programs (though often requiring state authorization) as they push toward Vision Zero goals.
Roundabouts: Safer Intersections by Design
Intersections are traditionally the most crash-prone spots in any network – it’s where paths cross and high-speed angle collisions (T-bones) or left-turn conflicts can occur. Roundabouts offer a redesign that virtually eliminates the deadliest intersection crashes. In a roundabout, traffic flows in a circle around a central island, with all vehicles traveling in the same direction and at lower speeds. There are no red lights to race through – every car must yield on entry and navigate a gentle curve, which naturally keeps speeds around 15–25 mph. The geometry also means any collision that does happen is a low-speed sideswipe rather than a head-on or T-bone. The safety impact of converting a traditional intersection (with signals or stop signs) into a roundabout is well documented: studies by the Insurance Institute for Highway Safety found about a 90% reduction in fatal crashes and 75% reduction in injury crashes on average. Even pedestrian risk goes down by about 30–40% because pedestrians only cross one direction of traffic at a time (with a refuge island halfway) and drivers are moving slower and more predictably.
Real-world example: Carmel, Indiana is known as the “Roundabout Capital” of the U.S. for its aggressive adoption of this measure. Carmel (pop. ~100,000) replaced dozens of four-way stops and signals with over 120 roundabouts citywide over the past two decades. The results are striking: back in 1996, Carmel had about 30,000 residents and 217 crashes that year. In 2019, with three times the population, the city saw fewer than 200 crashes total. Serious and fatal crashes have become extremely rare. The design has literally engineered out the most dangerous conflicts – an illustration of how forgiving geometry can save lives. Carmel’s success has inspired cities from Golden, Colorado to Kansas City, Missouri to construct more roundabouts in place of problematic intersections. Even rural areas are converting stop sign crossroads to roundabouts to cut down on deadly right-angle collisions. Of course, educating drivers on how to use roundabouts is key (navigating them was new to many Americans). But once accustomed, communities almost always report not just safety benefits but smoother traffic flow and less idling. In short, roundabouts offer that rare win-win: fewer crashes and less congestion. No wonder the American Association of State Highway officials recommends them as a default intersection type where feasible.
Data-Driven Decision Making: The Urban SDK Approach
At Urban SDK, our intuitive reports are empowering transportation departments to make faster, smarter, and more effective decisions. We ingest data from connected vehicles and turn it into clear, actionable insights for our customers.
Our geospatial AI tools at Urban SDK allow cities to rank roads by risk, monitor trends in crashes and speeding, and evaluate the impact of each safety intervention in near real time. For example, when Camas implemented our platform, they discovered that a handful of intersections accounted for a disproportionate share of traffic issues. This allowed them to install targeted solutions — and quickly measure results.
Our platform doesn't just guide internal decisions — it also helps city leaders build public trust. By sharing real-time safety data through open data portals and interactive maps, cities can communicate the “why” behind each decision. When residents see hard numbers — like speeding patterns on their own street — they're far more likely to support enforcement or street redesigns.
Key Takeaways for Safer City Streets
City planners and transportation officials can draw several important lessons from these 2025 trends:
- Blend High-Tech and Low-Tech Solutions: The strongest road safety gains come from combining AI tools with proven roadway design. Smart signals and analytics amplify the impact of traditional fixes like better crosswalks and lighting – it’s not either/or but both.
- Prioritize Vulnerable Road Users: Whether deploying AI or building roundabouts, always consider pedestrians and cyclists. Technologies like connected vehicles should incorporate pedestrian apps or detection, and street designs must slow cars down where people walk and bike. A safer city is one where the most vulnerable are protected by default.
- Use Data to Drive Decisions: Leverage data platforms like Urban SDK to identify problem areas and measure what works. An evidence-based approach ensures you focus investments on the interventions that will save the most lives. It also helps win grants and public trust by showing concrete facts and results.
- Speed Management is Key: Almost every success story – from Hoboken’s zero deaths to NYC’s injury reductions – involved slowing down traffic. Tools like speed cameras, raised crosswalks, and lower speed limits are extremely effective at taming speeds. Small reductions in speed make a huge difference in survivability, so treat speeding as the critical danger it is.
- Invest in Safe Intersection Design: Intersections should be designed for fail-safe outcomes. Where feasible, install roundabouts or add features like dedicated turn phases, improved signage, and high-visibility crosswalks. Reducing conflict points and enforcing yielding (as roundabouts do) can nearly eliminate deadly intersection crashes.
At Urban SDK, we’re proud to support cities across the U.S. in using real-time data and geospatial AI to reduce crashes, slow down traffic, and save lives. Our platform is helping shape the future of smarter, safer streets.
By embracing innovation while sticking to the fundamentals of traffic safety, U.S. cities are making strides toward the Vision Zero goal of zero traffic deaths. The year 2025 finds us at an encouraging crossroads: new AI technologies are helping anticipate and prevent crashes, and decades of road safety wisdom are being applied more broadly than ever. City leaders who harness both – the power of data and the power of smart street design – are delivering tangible results in the form of safer, more livable communities. It’s an exciting time for road safety, and the best practices emerging now can serve as a blueprint for cities everywhere to save lives on their streets.

TRAFFIC ENFORCEMENT FEATURES
80% of citizen complaints
are a perception problem
Urban SDK provides precise hourly speed data to evaluate complaints and deploy resources efficiently for the greatest impact to public safety.
Urban SDK provides precise hourly speed data to evaluate complaints and deploy resources efficiently for the greatest impact to public safety.
Target Speeding
Identify hot spots, validate monthly speeding trends and monitor vulnerable areas like school zones.
Improve Safety
Crash and citations location information to compare speed trends month over month
Fast Response
Respond to citizen complaints sooner with address search and exportable reporting
Deploy Assets
Generate maps for traffic enforcement by time of day, location or division to deploy officers to known problem areas.
RESOURCES
Customer Success
See how public sector leaders succeed with Urban SDK.
WEBINAR
Identify speeding and proactively enforce issues
See just how quick and easy it is to identify speeding, address complaints, and deploy officers.
