From Xi’an to Bangkok: AI Solutions Transforming City Traffic for the Better

It doesn’t take long for a driver stuck in a traffic jam in an Asian country to realize the city isn’t getting bigger. If anything, it’s becoming more overwhelming by the day.

With population rising and shifting towards the city, the concept of ‘complete streets’ is finding a resurgence amongst city planners in recent years—inclusive and safe pathways designed for a diverse range of users. For cities embracing this concept, it quickly becomes clear ‘complete streets’ is a multifaceted challenge of city planning, requiring the collaboration of various disciplines, from city planners and traffic departments to landlords.

While acknowledging that this transformation is a journey, the integration of big data collection through networks of surveillance cameras and AI technologies offers solutions that are unprecedented; helping cities rethink their roads. Here, we present three compelling data and AI use cases from densely populated areas that have significantly contributed to the creation of more livable societies by alleviating traffic congestion.

Anticipating congestions using various AI models in China’s Xi’an city increases traffic throughput

Due to Xi’an’s ancient architectural walls, road expansions are restricted, posing a challenge for its growing population of 2 million vehicles. To address this, the city has turned to data and AI to reimagine traffic management. One notable solution involves using traffic data to predict congestion.

By collecting data through cameras and sensors, the system learns about traffic patterns and can adapt rules accordingly. AI enables the development of various algorithms to manage congestion and improve traffic flow. The goal is to anticipate congestion before it occurs, implementing proactive solutions. Additionally, the model enhances incident prevention and road safety.

During its initial experimental phase documented in an article from Tomorrow City, the project has led to a 10% improvement in traffic flow and a 12% reduction in average urban journey times. The effectiveness of incident detection has also increased. However, despite advancements in AI and the potential of self-driving cars, completely eradicating traffic jams remains challenging. The physical constraints of cities like Xi’an, with over two million vehicles, continue to pose significant obstacles. Urban centers must acknowledge the need for a shift in traditional transport models to address these ongoing challenges.

Xi'an's architectural background creates an interesting challenge for traffic management

AI-Based Parking Enforcement in Bangkok's Ratchaprasong Project: A Solution for Congestion

In bustling urban environments with narrow and long main roads like Bangkok, managing illegal parking is a critical aspect of mitigating traffic congestion and ensuring the smooth flow of vehicles. With limited space and a growing number of vehicles on the road, unauthorized parking can lead to significant disruptions and delays for both motorists and public transport systems.

To address this challenge, authorities are leveraging technology and collaboration to enforce parking regulations effectively. Utilizing footage from CCTV cameras to capture vehicles' license plates, authorities can identify and penalize offenders efficiently. This proactive approach not only discourages illegal parking but also streamlines enforcement efforts, reducing the need for constant police intervention. "We cannot send police to chase off illegally parked vehicles all the time," said Bangkok deputy governor Wisanu Subsompon in an interview with Bangkok Post.

Furthermore, partnerships between public transport operators and local businesses are instrumental in providing alternative parking solutions. By establishing designated parking spots, drop-off locations, and waiting areas for passengers, stakeholders aim to alleviate congestion hotspots and enhance the overall traffic flow in the area.

Illegal parking creates traffic bottlenecks in Bangkok

The Circles Consortium experiment in Nashville showed how just one car with AI deployed effectively reduced jams

During Thanksgiving travel, countless individuals will experience sudden traffic standstills on interstates, often without any discernible cause such as construction or accidents. Researchers attribute this phenomenon to ‘Phantom Jams’ - the accumulation of abrupt brakings and sudden accelerations in navigating dense traffic conditions.

However, a recent experiment conducted using artificial intelligence in Nashville offers promising solutions to this issue.

Researchers from the Circles Consortium, a collaborative effort involving scholars from Vanderbilt University and several other universities, partnered with industry leaders including Nissan North America, Toyota, GM, and the Tennessee Department of Transportation to conduct an experiment. Their objective was to investigate how AI-equipped cars could mitigate phantom jams and improve traffic flow. By simulating real-world driving conditions and leveraging the expertise of multiple stakeholders, the study aimed to assess the potential impact of AI technology in reducing traffic congestion and enhancing overall road performance.

According to an abstract on The Vanderbilt University News portal, the test involved 20 cars on a closed track, with just one vehicle equipped with an AI system—combining adaptive cruise control with AI modifications to react to both nearby and distant traffic conditions. Remarkably, this single AI-equipped car changed the driving behavior of the entire fleet, alleviating the stop-and-go pattern that often leads to unexplained traffic jams. The ripple effects of this intervention resulted in significant fuel savings compared to typical traffic scenarios.

Furthermore, in an article with Fox, researcher Daniel Work highlighted the benefits of reduced 'stop-and-go' driving. This included a staggering 98% reduction in braking, a 40% improvement in fuel efficiency, and a 14% increase in distance traveled.

AI as a co-creator for complete streets?

As urban populations are projected to swell, with estimates by the UN suggesting that 70% of the world’s population will inhabit city spaces by 2050, dealing with traffic becomes an ever-pressing priority.

The examples showcased above vividly illustrate the transformative potential of AI in addressing the complex issue of urban traffic congestion. From accurately predicting jams in Xi’an to efficiently enforcing parking regulations in Bangkok, and even positively influencing driving behavior in Nashville, these innovative solutions are reshaping how we navigate our cities. However, addressing traffic congestion is no simple feat—it's a challenge that involves calibrating physical infrastructure as well as negotiating human behavior.

Yet, as we peer into the future, it becomes evident that AI will assume an increasingly crucial role in supporting traffic reduction and, perhaps eventually, in co-creating more livable urban environments starting from our streets.


TICTAG is a leading provider of data collection, annotation, and AI enhancement services, dedicated to revolutionizing industries through the power of artificial intelligence. Specializing in maximizing AI-enabled ROI, TICTAG empowers organizations to unlock new levels of efficiency and innovation. With a focus on enabling organizations to harness the full potential of AI, TICTAG offers cutting-edge solutions tailored to meet the unique needs of clients across various sectors. By combining advanced technologies with unparalleled expertise, TICTAG is committed to driving innovation, enhancing efficiency, and unlocking new opportunities for businesses worldwide.

With offices strategically located in Singapore, South Korea, India, Hong Kong, Indonesia, and Malaysia, TICTAG operates globally to cater to the diverse needs of its clients across different regions.

For more information, visit TICTAG's website.