Development and Implementation of an Intelligent Traffic Signal Control System (China)
Project Overview
With the rapid growth of China’s car ownership in recent years, traffic congestion has become a serious problem in many Chinese cities. Traffic signal control is an important means to manage urban traffic, and a “smarter” traffic signal control system can mitigate traffic congestions. However, in China, frequent road construction projects and fast-growing car ownership increase the instability of traffic patterns, which poses great challenges to effective traffic signal control operations.
Pretimed signal control systems are currently prevalent in China. Signal timing plans are usually updated based on results from traffic surveys every 1-3 years. These methods, however, provide no comprehensive or long-term traffic data, and traffic trends cannot be tracked in real-time. These pretimed signals consequently are unable to adapt to the rapid changes in traffic patterns, and any timing plans deployed can quickly become inefficient. While there are some applications of adaptive signal control systems, such as SCOOT, SCATS, in China the performance of these systems is often not ideal. There are two main reasons for this, first, these systems are highly dependent on data inputs from loop detectors, which are often easily damaged in China. Second, systems like SCOOT and SCATS are based on traffic models modelling traffic participates’ behavior in western countries, but Chinese traffic participants’ behaviors are quite different. However, recent advances in detection and communication technologies have enabled multidimensional traffic data collection in real-time. Developing and implementing a new intelligent signal control system which is truly adaptive at a relatively low cost is more practical than ever.
This collaborative project with BSIC BigData aims to develop an intelligent signal control system specializing in solving Chinese traffic problems. The system collects real-time traffic data from multiple sources, such as through computer vision and Bluetooth and Wifi sniffing, and displays and analyzes the data using a big-data platform. By feeding the data into a machine learning algorithm, the system can learn how to adapt the signal timing plans to current traffic conditions by itself, and improves its performance over time. The system is currently under field test in Huai’an, Jiangsu Province, and Zunyi, Guizhou Province, in China.
Project Partner
BSIC (Shanghai) BigData Technology CO., LTD (辉略（上海）大数据科技有限公司), founded on July 25th, 2017, is a technical company specializing in big data application platform development. Since 2017, BSIC BigData and iTSS Lab have forged a strategic partnership to promote technology transfer and innovation. Currently, BSIC BigData is developing and promoting big data platforms in the field of transportation (urban signal control optimization, traffic budget planning), environment (PM2.5 pollution detection and control), medical treatment (hospital Wi-Fi positioning, medical record matching), and automobile (user conversion rate modelling) in China.