For years, traffic data meant GPS probes and crowdsourced incident reports. But 2024-2025 marks a turning point as new data sources become commercially available. If you're building routing applications, these emerging sources offer opportunities to improve detection speed, coverage, and accuracy.
Edge AI Video Processing
The biggest shift in traffic data is the maturation of edge AI for video processing. Previously, analyzing video from cameras or dashcams required expensive cloud infrastructure with latency-inducing uploads. New edge processors can run sophisticated detection models directly on camera hardware.
What's New
Edge AI chips like NVIDIA Jetson and Qualcomm Snapdragon can now run traffic detection models that match cloud performance. This enables real-time incident detection at the camera or dashcam level, with only event metadata transmitted.
For data engineers, this means:
- Sub-second detection latency from visual sources
- Reduced bandwidth costs (events vs. video streams)
- Privacy-preserving analysis (video processed locally)
- Scale that was previously cost-prohibitive
Expanded Public Sensor APIs
State DOTs are increasingly publishing real-time sensor data through open APIs. What was previously siloed in traffic management centers is now accessible to developers. New developments include:
Real-Time Sensor Feeds
More states are publishing loop detector, radar, and weather station data through standardized APIs. California's PeMS system, Texas DOT's open data portal, and similar initiatives provide ground-truth traffic measurements.
Work Zone Data Exchange (WZDx)
The USDOT's Work Zone Data Exchange standard is gaining adoption. This provides structured data about construction zones, lane closures, and work zone speeds in a consistent format across states.
Smart Traffic Signal Data
Connected traffic signal systems are beginning to share timing data. This enables applications to factor signal phases into routing decisions and provide more accurate ETAs through signalized corridors.
Fleet Telematics Partnerships
Commercial telematics providers are increasingly willing to license aggregated data for traffic intelligence. Major developments:
- Delivery fleet data: Amazon, FedEx, UPS fleet data becoming accessible through partnerships
- Rideshare data: Uber and Lyft driver data available through official programs
- Trucking telematics: ELD mandate created standardized data from commercial vehicles
Coverage Expansion
Fleet telematics data now covers over 15% of vehicle miles traveled in the U.S., up from under 5% five years ago. This growth is driven by ELD mandates, insurance telematics programs, and fleet management adoption.
911 CAD System Integrations
Computer-Aided Dispatch (CAD) systems used by 911 centers are increasingly offering data sharing capabilities. NextGen 911 initiatives are standardizing interfaces that enable third-party access to incident data.
New capabilities include:
- Real-time incident notifications via API
- Structured incident classification (APCO/NENA standards)
- Unit dispatch and response status
- Estimated clearance times
Connected Vehicle V2X Data
While full V2X (vehicle-to-everything) infrastructure is still years away, early deployments are generating useful data:
- Signal Phase and Timing (SPaT): Connected signals broadcasting timing data
- Basic Safety Messages (BSMs): Early connected vehicles sharing position and motion
- Road weather data: Connected vehicles with ambient sensors
Current V2X coverage is limited to pilot corridors, but these represent new data types that will scale as connected infrastructure expands.
Aggregated Traffic Data APIs
Perhaps the most significant development is the emergence of platforms that aggregate multiple new sources into unified APIs. Rather than building integrations with each source individually, developers can access combined intelligence through a single endpoint.
These platforms handle:
- Source integration and normalization
- Deduplication across sources
- Confidence scoring and source attribution
- Reliability monitoring and failover
Key Takeaway
The traffic data landscape in 2024-2025 is fundamentally different from five years ago. Edge AI enables real-time video processing, public APIs are opening government sensor data, fleet telematics coverage is expanding, and 911 systems are becoming accessible. For routing applications, these new sources offer improvements in detection speed, coverage, and context that weren't possible before.
Published by
Argus AI Team