See traffic before it happens.
TrafficSense forecasts traffic flow across road networks using a globally aware spatiotemporal model, helping cities time signals, manage congestion and inform drivers.
The problem
Congestion costs time, fuel and safety.
Traffic depends on complex patterns across space and time. Conventional CNN-based models struggle to capture these dependencies and are slow when stacked deep.
real-world traffic datasets where our model achieved state-of-the-art results.
stacked blocks are enough for strong performance, keeping the model efficient.
attention modules capture both global and local spatiotemporal patterns.
Source: the founding research paper (IEEE Transactions on Intelligent Transportation Systems, 2025).
How it works
How it works
A clear path from data to decision.
Collect
Traffic sensor and detector data are gathered across the network.
Model
ST-WaveMLP learns spatial and temporal dependencies.
Forecast
Flow is predicted for each road segment ahead of time.
Manage
Operators adjust signals, routes and alerts.
Product
What makes it different.
Every feature is grounded in the founding research.
Global awareness
SGAC captures global spatial and local temporal dependencies.
Local precision
SLAC captures local spatial and global temporal dependencies.
Efficient
MLP-based blocks converge faster than stacked CNNs.
State of the art
Top results on four real-world datasets.
Who it's for
Who it's for.
Traffic management centres
Proactive signal and incident management.
City planners
Data-driven road and transit planning.
Navigation & logistics apps
Better arrival-time predictions.
The research
Built on peer-reviewed research.
- Proposes ST-WaveMLP, a globally aware spatiotemporal network for traffic flow prediction.
- Repeatable ST-WaveBlock with SGAC and SLAC attention components.
- Effective modelling with only 2–4 stacked blocks.
- State-of-the-art performance on four real-world traffic datasets.
Byeon, H., et al. (2025). Spatiotemporal Network for Accurate Traffic Flow Prediction in Intelligent Transportation Systems Using Generative AI. IEEE Transactions on Intelligent Transportation Systems.
Team
The team behind it.
Founded in 2025 by a student team from the faculty below.
Faculty of Engineering (Artificial Intelligence Engineering)
Spatiotemporal deep learning and traffic analytics.
FAQ
Questions we hear often.
What data does it need?
Traffic flow data from loop detectors, cameras or other road sensors.
How far ahead can it predict?
Prediction horizons are configured per deployment and dataset.
Is this a finished commercial product?
It is a research-based student startup moving from validated prototype toward pilots with real users.
Can we partner or run a pilot?
Yes. We are looking for organisations willing to test the solution in real conditions and share feedback.
Bring TrafficSense to your organisation.
We are onboarding pilot partners now. Read the research and get in touch.