AI Traffic Management Platform Reducing Peak Congestion 31% Across a City Centre
Challenge
Valeton’s city centre faced severe congestion, costing an estimated £87M annually. Traffic signals operated on fixed timing plans from 2009, with 340 junctions functioning independently and no real-time coordination. Air quality in two zones exceeded legal limits, creating regulatory pressure to act within 18 months.
Solution
We deployed an AI-driven adaptive traffic management system, integrating real-time sensor infrastructure, network-wide signal coordination, predictive optimisation, and public transport prioritisation.
Results
Peak congestion reduced by 31%. Bus punctuality improved from 64% to 84%. NO₂ levels dropped by 19%, bringing both non-compliant zones within legal limits. Emergency response times improved by 22%.
Valeton’s traffic system was designed for a different city. Signal timing plans last updated in 2009 were still in operation, despite major changes in travel patterns: new residential developments, shifting retail activity, increased micro-mobility usage, and a relocated hospital altering traffic flows.
340 junctions operated independently, with no coordination and no ability to respond to real-time demand. The result was predictable:
- peak congestion across key corridors
- inefficient routing across the network
- worsening air quality in high-density zones
At the same time, regulatory pressure was mounting. The council had 18 months to bring two zones into legal air quality compliance.
Building the Real-Time Data Layer
The foundation of adaptive traffic management is continuous, high-resolution data. We deployed a multi-layered sensor network across the city centre:
- Inductive loop countersInstalled across 890 approach lanes, capturing per-lane vehicle counts and occupancy at 1-second intervals
- ANPR cameras186 cameras tracking journey times and origin-destination flows (with data anonymised at capture)
- Bus AVL integrationReal-time position data from public transport systems, enabling delay-aware prioritisation
- Environmental sensorsNO₂ and particulate monitoring across 24 roadside locations, feeding a live air quality index
This created a unified, real-time view of traffic demand, movement patterns, and environmental impact.
Network-Wide Signal Coordination
The core of the platform was a city-scale signal coordination system. All 340 junctions were connected into a single control layer, enabling:
- synchronised signal phases across arterial routes
- dynamic coordination based on real-time demand
- removal of isolated, fixed timing behaviour
This transformed the network from independent nodes into a coordinated system.
AI-Driven Signal Optimisation
On top of the coordination layer, we implemented an adaptive optimisation engine based on SCOOT-derived algorithms enhanced with machine learning.
Key capabilities included:
- Rolling demand predictionForecasting traffic volumes 15 minutes ahead at each junction approach
- Dynamic phase adjustmentSignal timings recalibrated continuously based on predicted and live conditions
- Green wave optimisationProgressive signal coordination allowing vehicle platoons to move without stopping
- Incident detection and responseAutomatic plan switching when upstream data indicates disruption
- Bus priority logicConditional early-green or phase extension for buses running 3+ minutes behind schedule
The system operates continuously, recalculating and adjusting in near real time.
Air Quality Control Mode
Beyond congestion, the system directly addressed environmental compliance. When NO₂ levels exceed 75% of legal thresholds, a dedicated low-emission mode activates:
- longer signal cycles to reduce stop-start driving
- minimisation of idling at junctions
- diversion of through-traffic to outer routes
This mode has been triggered 23 times since deployment, delivering an average 14% NO₂ reduction during activation periods.
Measured Impact
The transition to adaptive control produced measurable outcomes across mobility, environment, and public services:
- Peak journey times: reduced by 31%
- Bus punctuality: 64% → 84% across 12 priority routes
- NO₂ levels: reduced by 19%, achieving legal compliance
- Emergency response times: improved by 22%
The system delivered both operational efficiency and regulatory compliance.
Why It Worked
The impact came from integrating three layers into a single system:
- Continuous sensingReal-time visibility across the network
- Centralised coordinationJunctions operating as a system, not isolated units
- Adaptive intelligenceAI-driven optimisation responding to changing demand
This shifted traffic management from static planning to dynamic control.
Final Thought
Urban traffic is not a fixed problem. It is a continuously changing system that requires continuous adaptation. Cities that rely on static infrastructure fall behind. Cities that build adaptive systems gain efficiency, compliance, and resilience.
Building Smart City Platforms?
Intagleo Systems helps governments and urban operators design real-time infrastructure platforms, optimise city-scale systems, and deliver measurable improvements in mobility and sustainability.