AI-Driven Smart Intersection Management

🚲 Mobility

📍 Istanbul, Türkiye / Helsinki, Finland

About

Status:

✅ Finalized

Istanbul is Türkiye’s largest city and a unique metropolis spanning both Europe and Asia across the Bosphorus Strait. With over 15 million residents, it is a major economic, cultural, and logistics hub connecting continents. The city combines historical heritage with rapid urban growth and infrastructure development. Istanbul is actively investing in smart mobility, disaster resilience, and sustainable urban services to manage its scale and strategic position.

🗓️ Start: 2026-03-25

⇨

🗓️ End: 2026-03-25

In major hubs like Istanbul and Helsinki, intersections are critical bottlenecks causing severe congestion and safety risks. Currently, monitoring is largely manual and fragmented, leading to slow response times. The AI4LIFE challenge implements an AI-based management system that integrates IoT infrastructure with real-time analytics. By automating the detection of abnormal traffic situations and consolidating multiple tools into a single hypervisor, the solution reduces response times, lowers emissions from idling, and enhances road safety for millions of commuters.

Description of the Challenge

Cities generate approximately 70% of global greenhouse gas emissions, with road transportation accounting for a quarter of this total. In Europe, traffic congestion costs €110 billion annually. Safety is a critical concern: 20,000 people die annually on European roads, with 50% of these fatalities occurring at intersections.

Local transformation context:

Istanbul: A metropolis of 18 million people where 97,354 traffic accidents were recorded in 2024. It faces the highest congestion levels in Europe, with an average of 105 hours lost per year per citizen.
Helsinki: The Jätkäsaari district (Europe’s busiest passenger port) acts as a geographic bottleneck where ferry arrivals unload hundreds of vehicles simultaneously into a dense urban grid.

Specific problem

How it manifests: Intersections are the primary “hotspots” for both traffic accidents and infrastructure bottlenecks. In Istanbul, the sheer volume of 34 million daily trips makes manual oversight impossible to scale. In Helsinki, the surge of vehicles from ferries creates immediate gridlock that clashes with public transport priorities.

Who is affected: Traffic management operators (suffering from cognitive overload), citizens (impacted by air pollution and lost time), and vulnerable road users.

Current situation and gap

Current approach: Traffic monitoring remains heavily manual. Operators must monitor camera feeds and intensity maps or wait for citizen phone calls to identify incidents.

The Gap: Monitoring tools are fragmented. Operators must juggle separate systems for data analysis, camera surveillance, and event tracking. This lack of integration prevents a proactive, real-time response to “abnormal situations” before they escalate into major accidents or total gridlock.

Expected Outcomes

The city needs an automated way to detect abnormal traffic patterns and a unified interface that consolidates fragmented data sources (cameras, sensors, mobile apps) to support real-time interventions.

Expected outcome

Desired solution: An integrated AI-based intersection management system (Software + IoT).
Functional expectations: Automation of incident detection, predictive traffic pattern analysis, and optimization of signal timings.
Visualizations: A single hypervisor dashboard (Kentyou Eye) providing real-time situational awareness and historical impact monitoring.
Expected impact:
- Environmental: Reduced CO2 emissions through optimized traffic flow.
- Social: Significant reduction in accident risks at critical junctions.
- Operational: Faster response times and streamlined workflows for traffic control centers.

Space for Solutions and Experimentation

Available experimentation space:
- Istanbul: Initially 4 major intersections, now scaled to 300 intersections monitoring over 1 million vehicles per day.
- Helsinki: 15 intersections in the Jätkäsaari district.
Scale of experimentation: Metropolitan (Istanbul) and District (Helsinki).
Available data/tools: 1,000+ data sources (cameras, sensors, TomTom data, mobile apps). Infrastructure provided by ISBAK (Istanbul’s municipal company).
Existing experience: The pilot demonstrated a “plug-and-play” capability, connecting to existing city infrastructure in just 1 to 2 days. It builds on previous award-winning research with Columbia University.