Bangkok Cut Delays by 24% with Gridsmart® AI Signals

Overview

Bangkok is one of the world’s most vibrant cities—alive with energy, people, and movement. But for years, its traffic told a different story: gridlocked intersections, unpredictable delays, and a system built for a fraction of today’s demand.


Originally designed to accommodate around 2 million vehicles, Bangkok’s roads now handle over 8.5 million every day. Combine that with the world’s second-highest rate of motorbike ownership along with a vibrant tourism economy, and the need for modern traffic management becomes crystal clear.

For too long, Bangkok’s infrastructure relied on fixed-time traffic signals—rigid systems that couldn’t adapt to real-time flow.

That’s when Thailand’s Ministry of Transport (MOT), in partnership with New Trend Development and Cubic’s Gridsmart® video detection, set out to rethink how intersections should work in the 21st century.

Their goal wasn’t just to reduce congestion. It was to build a smart, safe, and scalable traffic management system that would work with the city; not against it.

 

The Problem: Outdated Tools in a Rapidly Growing City

While Bangkok’s population and vehicle count skyrocketed, its traffic systems stood still. The city continued to depend on fixed-timing signals, pre-set light cycles that didn’t account for real-time conditions or dynamic traffic behavior.

The MOT’s initial effort to modernize included testing inductive loop detectors, but those solutions quickly proved unfit for Bangkok’s environment due to:

• High maintenance costs: Constant wear from Bangkok’s heavy use made upkeep time-consuming and expensive.
• Unreliable accuracy: During high-traffic periods, sensors struggled to deliver clean data.
• Missed motorbikes: In a city where motorbikes make up a huge portion of daily traffic, failing to detect them created major blind spots.

Without complete, real-time data, adaptive traffic systems couldn’t function effectively, resulting in unnecessary delays, growing queues, and increased accident risk.
 

Bangkok needed a more adaptive intersection solution—one that could account for all road users and constantly changing conditions.

 

The Solution: Gridsmart Intersections Powered by AI + Computer Vision

To solve these challenges, Bangkok implemented Gridsmart, an intelligent intersection solution built on AI-driven video detection, real-time object classification, and dynamic signal actuation.

A Smart View from Above

Rather than relying on unreliable in-pavement sensors, Gridsmart leverages a single 360° fisheye camera mounted above each intersection. This unobstructed view enables the system to:

• Detect and classify all road users in real time—from cars and buses to motorbikes, bicycles, and pedestrians
• Track their movement through the entire intersection, not just at stop lines
• Focus on the center of the intersection, where conflicts and collisions are most likely to occur

With more comprehensive visibility, the system captures the nuance of Bangkok’s traffic flow, including the high volume of motorbikes that previously went undetected.

Real-Time Signal Adjustments, Driven by Data

The insights collected from each intersection are streamed directly into Thailand’s adaptive signal control platform via Gridsmart’s open API, enabling real-time signal changes based on current demand.

This live feedback loop allows traffic signals to:

• Adjust cycle lengths dynamically

• Prioritize heavily trafficked lanes during peak times

• Minimize unnecessary idling during off-peak hours

The results were immediate and undeniable. In the pilot program across three major intersections, delays dropped by up to 24.52%, and average queue lengths shrank by 30.5%, leading to more predictable, efficient commutes for thousands of daily drivers.

 

Perhaps the most impressive of all: Gridsmart’s deployment resulted in a projected $855,000 in annual cost savings, combining reduced fuel use, shorter travel times, and fewer delays.

What began as a smart infrastructure upgrade proved to be a high-return investment in Bangkok’s long-term safety, mobility, and sustainability.

Easy to Maintain, Built to Scale

Because Gridsmart operates above ground, it bypasses many of the maintenance challenges faced by traditional solutions. Its design is resilient in Bangkok’s variable weather, and its modular architecture allows for gradual rollout—intersection by intersection, not all at once—making it a scalable and budget-conscious solution for cities of all sizes.

 

Key Results

Third-party testing of the Gridsmart deployment across three major intersections in Bangkok revealed powerful outcomes:

• Delays decreased by up to 24.52%. Intersections responded in real time, reducing stop-and-go inefficiencies.
• Queue lengths cut by as much as 30.5%. Reduced backups made for less congestion and faster passage.
• Red-light running dropped by up to 68%. Adaptive signals + clearer visibility encouraged safer driving behavior.
• 53K commuter hours saved annually; real time given back to residents, businesses, and the economy.
• $855,000 in projected annual savings. Cost savings came from reduced fuel consumption, shorter delays, and fewer traffic incidents.

 

Conclusion

Bangkok’s success was about easing congestion and proving what’s possible when cities adopt intelligent, forward-thinking solutions.

With Gridsmart, the city upgraded strategically, without tearing up roads or starting from scratch. It tapped into the power of AI, real-time vision, and modular deployment to create an infrastructure that is:

• Scalable for citywide expansion
• Inclusive of all road users
• Built for today and ready for tomorrow

A smart intersection solution doesn’t just move traffic. It moves your city forward.