next-gen traffic lights powered by AI & state of the art multi-agent reinforcement learning algorithms

trafluid

Real-time, cycle-free, AI-powered adaptive traffic signal control.

State of the art reinforcement learning techniques

Unlike traditional adaptive systems that operate on fixed cycles, Trafluid leverages advanced reinforcement learning to dynamically activate the most efficient signal phase at any moment. By analyzing both real-time and historical traffic data, our AI models learn optimal control strategies and respond instantly to changing conditions across the entire road network. Integrated directly with traffic signal controllers and cabinet equipment, Trafluid ensures every decision is not only efficient but also 100% safe.

Watch Video

Cut Commute Times and Costs with Trafluid

Trafluid reduces commute times by up to 40% compared to fixed-time signals, and by over 10% even when measured against optimized, coordinated systems. Best of all, these performance gains come without the usual price tag: Trafluid is up to 50% more affordable than traditional solutions, thanks to its cloud-based design and hardware efficiency!

Request Demo

vtl smartsignal

Fully automated AI-powered signal retiming

Dynamic Optimization

Continuous AI-driven adjustments to signal timing plans based on observed and predictive traffic data - Tailored updates at intervals chosen by the municipality—every 5 minutes, hourly, daily, or otherwise.

Download Whitepaper

Automated Deployment

Automated deployment of optimized timing plans to controllers via SNMP over the NTCIP 1202 standard - Eliminates manual interventions, reducing staff workload and human error.

Request Demo

vtl smartsense

AI-based multi-modal vehicle detection system

Accurate Real-Time Detection at low cost!

SmartSense can detect vehicles at all approaches with >90% accuracy and just a fraction of a second of delay, while being 50% cheaper than other solutions on the market. SmartSense is compatible with existing controllers and has minimal hardware requirements!

Request Demo

vtl priority

Prioritization system for any type of vehicle

Prioritization on demand

VTL's Prioritization System seamlessly integrates with existing traffic signal control systems to prioritize any vehicle using a simple smartphone app. While ideal for emergency vehicles, it also enables prioritization for commercial fleets, opening new opportunities for businesses and mobility services.

Request Demo

vtl digital twin

Recreate Real Traffic with High-Fidelity Simulation

High-Fidelity Simulation

VTL Digital Twin replicates your area of interest with exceptional detail, using historical traffic data to accurately simulate real-world traffic flows. Test traffic signal strategies and infrastructure changes in a fully virtual, risk-free environment. Evaluate performance with precision across key metrics, saving real-world testing for only the best-performing scenarios. VTL Digital Twin supports road networks with over 1,000 intersections and can run simulations 10x to 100x faster than real time, depending on network size and traffic density.

Request Demo

Download Whitepaper

Learn more on VTL SmartSignal retiming solution.

OUR Market

Municipalities & Cities

State and local Departerments of Transportation (DOTS)

Connected Autonomous Vehicles

OEMs, Tier 1s, and Tier 2s, and other automotive companies

Developing Countries

Countries with limited infrastructure and resources

Controlled Spaces

Airports, seaports, warehouses, and other controlled spaces

Smart Cities

Cities investing in advanced infrastructure and technology

Transportation Companies

Trucking, shipping, and other transportation companies

About Us

VTL is a startup spun out of Carnegie Mellon University (CMU) where extensive research was conducted. We have licensed 20+ patents in the fields of AI, V2V, V2X in Traffic Management.

Ozan Tonguz

CEO & Founder

Inci Ozgunes

Co-CEO & Co-Founder

Khaled Al-Ghoneim

Co-Founder

Vijayakumar Bhagavatula (Kumar)

Co-Founder

Abdou Lazraq

Head of Engineering

Federico Taschin

Machine Learning Engineer

Our Partners

Our partner network is growing and we are always looking for new partners to collaborate with.

Get In Touch

Open to all opportunities and collaborations. Please feel free to contact us.

US Office Address:
5503 Covode St. Pittsburgh, PA 15217, United States

EU Office Address:
Ivana Kukuljevica 10 42000 Varazdin Croatia, EU