They say data is the new oil, but it’s only valuable if you can refine it into something useful. In transportation, that means turning raw traffic data, sensor inputs and video feeds into real-time insights that actually improve how we move.
That metaphor set the tone for a 2025 IoT Slam panel focused on how artificial intelligence and the Internet of Things (IoT) are reshaping one of society’s most universal pain points: traffic.
With the average American spending 51 hours stuck in congestion each year, and the broader economy losing an estimated $90 billion annually because of it, the urgency for innovation is clear.
But beyond congestion, the panel explored broader themes like digital infrastructure, predictive maintenance, equitable AI adoption and ultimately, the potential to save lives.
1. Don’t just collect data, design for decisions
Kelly Wells, Traveler Information Engineer at the NC Department of Transportation (NCDOT), oversees traffic management for more than 80,000 miles of roadway and over 1,000 traffic cameras. She knows firsthand how critical – and overwhelming – data can be.
“Data collection is only the start,” she said. “Our operators jump between three different systems just to manage crashes, cameras and highway signs. We’re in the painful process of centralizing those tools, but the hardest part is geospatial conflation, making all that data speak the same language.”
Her advice: collect strategically and integrate early. Not every asset needs a sensor, but the right ones, combined with AI, can lead to smarter, faster responses in high-stakes moments.
Why it matters: Turning data into real-time decisions can reduce emergency response time, increase system efficiency and save lives.
2. Make aging infrastructure smarter, not just newer
Elizabeth Young, VP of Asset Management and Data Analytics at Halff, emphasized that municipalities don’t need to rebuild from scratch. They need to retrofit what they have.
“Cities are often the last to adopt new tech,” she said. “But they’re sitting on valuable data. By monitoring key assets, like bridges and roadways, with sensors and analytics, they can extend lifespan and prevent failures.”
Young also pointed to metadata as a commonly overlooked key to successful integration. “Understanding where your data came from helps extract more value from it.”
Why it matters: Predictive maintenance through AI and IoT helps cities do more with less. It extends the life of assets while reducing the risk of failure and costly downtime.
3. Build connected systems, not digital silos
As agencies invest in new technology, Tyson Echentile, SAS Global Manager for IoT Partner and Business Development warned against replacing physical silos with digital ones.
“One department’s dashboard shouldn’t be another’s blind spot,” he said. “To unlock the full value of AI, agencies must adopt a systems-of-systems approach, integrating across departments, not just within them.”
Wells echoed that frustration. “You think everything’s integrated until someone asks how to combine crash data with speed data, and you realize they’re on two different referencing systems.”
That’s why it’s not enough to add new tech on top of old problems. According to SAS’ Katy Salamati, it starts with platform design:
“It’s very important to work with platforms that cover the whole data and AI lifecycle. From managing data quality and integrating various sources – real-time, demographics, address data – to building and operationalizing models that actually drive decision-making. You don’t just want to build models for the sake of it – you need a system that can manage and scale them in production.”
Why it matters: If “data is the new oil,” then agencies need a reliable refinery. A system-of-systems approach enables smarter, faster coordination across departments, turning raw information into real impact.
4. Design AI that taps people on the shoulder
Technology is only as good as its adoption. For Wells, the goal isn’t to overwhelm operators, it’s to empower them.
“I want systems that push information to people,” she said. “A tap on the shoulder: ‘Hey, this crash just happened, here’s the right camera feed and here’s a message suggestion for the sign.’ That’s far more useful than expecting someone to pull the right info from five screens.”
Salamati, who consults organizations about IoT, agrees. “The best AI systems don’t just analyze. They assist,” she said. “We need platforms that manage the full AI lifecycle, from data quality to model production, to help decision-makers in real time.”
Why it matters: Human-centered design ensures that AI becomes a copilot, not a hurdle, for public sector workers on the front lines.
5. Use real-world results to build public trust
One standout use case: SAS’ traffic signal optimization project in a major UAE city.
“We used AI to reduce traffic light delays along a busy corridor between the airport and downtown,” said Salamati. “It resulted in a 15 to 20 percent decrease in congestion.”
This isn’t just about convenience. It’s about safety.
“Every day, I read a fatal crash report,” said Wells. “We don’t call them accidents. They’re crashes, often preventable. And autonomous vehicles can help eliminate those causes.”
Why it matters: As cities like Los Angeles gear up for events such as the 2028 Summer Olympics, real-world success stories will help communities embrace AI’s potential, especially when the outcome is fewer crashes and safer roads.
Looking ahead: What excites transportation leaders most?
- Elizabeth Young sees enormous potential in digital twins, virtual environments where planners can simulate traffic scenarios before breaking ground.
- Katy Salamati is inspired by multimodal innovation, how AI is helping cities synchronize buses, rail and ride-sharing for seamless journeys.
- Kelly Wells believes the biggest win is safety. “I really believe that when the vehicles are autonomous, a lot of those things that contribute to these crashes will be taken away.”
