13 Sept 2016
Could Big Data Ease Traffic Congestion in Tomorrow's Smart Cities?
This year's Guangzhou International Smart City & Security and Protection Products Exhibition was dominated by big data systems, many of which have been designed to take traffic management and incident response to a whole new level.
The promise of an end to – or at least, an easing of – traffic congestion in a number of mainland cities was the talk of this year's Guangzhou International Smart City & Security and Protection Products Exhibition. Overall, those exhibitors promoting the use of big data as a way to improve urban traffic, as well as its use in a variety of other applications, were among the most sought out. The most eye-catching stands, however, featured a series of virtual reality systems capable of simulating the real world driving environment.
According to the National Development and Reform Commission, several innovative big data platforms will come online over the next two to three years, all geared to facilitating enhanced exploitation of the ever increasing volumes of information generated every day. According to research by the China Industrial Information Network, the global big data market was worth some RMB150 billion in 2015, a 24.2% year-on-year increase. Over the same period, China's big data market was valued at RMB16 billion. While accounting for only 10.7% of the global market, this represented a year-on-year growth of 65.3%, a figure 2.7 times the global average.
The prominence of big data at this year's exhibition was also seen as a clear indication of the importance China attaches to big data with regard to its smart city construction programme. A case in point is Hangzhou-based Yuantiao Tech, a company that specialises in applying big data analytics to highway statistics, providing smart traffic information services to government departments, businesses and the public.
Tong Wenzhe, the company's Key Account Manager, said: "Our big data analytics mainly focus on urban congestion control." Over on the company's stand, five screens displayed big data-related congestion control methods. Overall, the company's methodology is said to involve five primary steps – congestion location detection, status analysis, control solutions, congestion forecasting and public routing.
In the initial stages, the company collects data to assess traffic conditions. This is conducted via the use of traffic flow and volume monitoring systems, all set to collect data at the congestion location. While the smart traffic flow monitoring system acts to monitor road conditions in real time, the volume alert system then detects road sections with unusually high activity levels.
Congestion status analysis is then conducted using a traffic light-assisted decision system. This allows the signal timing of traffic lights to be assessed and analysed in order to get an understanding of the routes preferred by road users. Once this has been completed, the signal timing optimisation system can be used to devise congestion control solutions.
According to Tong, as the signal timing of traffic lights impacts the entire road network, using big data analytics to set signal timing can ease the traffic flow along key routes. Signal timing solutions, however, would not be applied to actual roads immediately, with the company first performing simulations and analysing the likely outcome before applying the most appropriate solutions in a real world situation.
Yuantiao also uses big data to conduct congestion forecasting and to predict the likely traffic congestion caused by any major incidents. For a traffic control department, using the company's products would give them advance warning of the likely traffic condition should such an incident occur. This would allow them to optimise their deployment of traffic control officers in a bid to minimise any such problems.
Overall, the key element of big data is the core algorithm. As Yuantiao specialises in big data road traffic analytics, the company believes it has a clear competitive advantage in the sector.
Eying another sector is Beijing DeepGlint Technology, with the company focussing on the demand for video big data applications. Over on its stand, staff members were keen to highlight its Weimu vehicle feature analysis system. This system can recognise more than 3,000 different vehicle models, 19 specific vehicle types and 12 colours. Its day-time recognition rate is above 95%, with night-time recognition falling to 90%, while a licence plate recognition accuracy in excess of 98% is maintained. By applying this technology to public security and using it for traffic management, vehicles with forged licence plates can easily be identified.
Internet+ represents a combination of internet information technology with more conventional approaches in order to create dynamic new systems. According to the latest Statistical Report on Internet Development in China – released by the China Internet Network Information Centre (CINIC) in December 2015 – last year China was home to 688 million internet users, while its internet penetration rate stood at 50.3%. This shows that more than half of China's population has internet access, with some 90.1% of these users going online via mobile devices.
One company keen to embrace the internet+ approach is Shenzhen Yunovo Technology, with its representatives keen to promote its smartcar rear-view mirror at the fair. According to Yan Song, the company's Key Account Director, the mirror is connected to the internet via the Android operating system, with users having the facility to use voice control to obtain routes from built-in map software, as well as to control music players and other accessories.
According to Yan, one of the most impressive features of the mirror is its facility to capture images of traffic incidents with just one click. The user can either take photographs using voice control or by pressing a Bluetooth-enabled button on the steering wheel. When a driver sees any vehicles violating traffic regulations, they can store the footage from 10 seconds before pressing the button and for up to five seconds after. This is then automatically uploaded to the traffic police department's traffic violation report platform.
Internet-based smart products can also help facilitate law enforcement by the traffic police. The Law Enforcement Video Recorder – launched at the exhibition by Suzhou Keda Technology – is a smart product set to replace the current conventional approach. With 4G internet connection and a link to a back-end management platform, it can locate police personnel, provide a live view of traffic control and deployment, while also offering real-time video surveillance.
At present, mobile technology not only maximises consumer convenience, it also enhances the credibility of big data analytics. According to Tong, the 'suggest a route' function is currently its most popular with consumers.
By downloading the company's app, users can identify just which road sections are congested and can then get alternative route suggestions. At the same time, the company's big data analytics system obtains data on traffic conditions from government departments as well as from other users.
Any motorists using Yuantiao's app automatically share data with the company. This allows more accurate information on real-time road conditions to be obtained.
Also at this year's exhibition, there was wide range of VR (virtual reality) products on offer. A large number of these exhibits utilised the system as part of a driver training programme, with a number of attendees testing out the VR goggles on offer as they accessed the driving simulators.
Shanghai Chenglian Traffic Intelligent Technology was just one of many showcasing a VR driving simulator at the event. Its product can simulate both a typical driver training ground as well as a more authentic urban driving environment. Any trainee driver can gain experience across variety of road conditions, helping them to master advanced skills in a wholly safe environment.
According to Liu Yuan, Chenglian's Representative at the event, the company can also tailor-make specific scenarios for individual customers. In the case of the Anhui province, for instance, where there are many hilly roads, a special module could be created to simulate these unusual conditions.
Using the same VR technology, the Traffic Management Research Institute of the Ministry of Public Security has developed a safe riding experiential system that can help build awareness of potential hazards. According to Yan Xilei, a Researcher at the Institute, people using this product can gain a better understanding of the many potential perils facing road users.
Elsewhere at the event, an expressway-sited traffic forensics system, developed by the Zhejiang-based Senken Group, was on show. This automatically collects information from any location where traffic violations have been committed. Similarly, a smart traffic direction management platform launched by Nanjing's Duolun Technology integrates real-time video footage and video recordings at key road junctions, while offering traffic direction, service direction, electronic policing and roadblock-setting functions.
The Guangzhou International Smart City & Security and Protection Products Exhibition 2016 was held at Pazhou's Poly World Trade Expo Centre in Guangzhou from 29-31 August.
Xing Bin, Special Correspondent, Guangzhou