Skip to main content

Hong Kong's integrated traffic management system

Hong Kong’s Route 8 now features an extensive and advanced traffic control and surveillance system developed to overcome challenges of great scale and complexity, write Delcan vice president Rex Lee and MD Joseph Lam
May 22, 2012 Read time: 6 mins
Route 8 is a complex stretch of road which demanded an innovative traffic control and surveillance system

Every day, over 11.6 million passenger journeys are travelled within the province of Hong Kong, with about 610,000 vehicles using 2,076 kilometres of roads. There are approximately 290 licensed vehicles for every kilometre of road. Managing traffic in Hong Kong is never easy.

Over the past 20 years, to maximise the capacity of the road network and enhance road safety, the city has deployed advanced technology for traffic management. The city started Intelligent Transportation Systems (ITS) with extensive deployment of adaptive traffic signal control, CCTV cameras, and tunnel traffic control and surveillance systems (TCSS). There are currently seven adaptive traffic signal control systems and over 14 TCSS in Hong Kong. In the late 1990s, use of TCSS was extended to area-wide traffic management for Hong Kong’s road network of a mix of highways, bridges and tunnels.

In recent years, the focus has shifted to integrated management of the city’s strategic road network, with deployment of traffic speed mapping, journey time indication and the latest initiative of planning for an area wide traffic incident management system.

This move from deployment of isolated systems to integrated traffic management and supply of network-wide traveller information represents the typical challenge that many agencies are facing today; addressing the need for interoperability and standardisation of data. The Route 8 TCSS is the latest to interface with adjacent traffic management systems and provides data links for traffic information integration.
 

The Route 8 Traffic Control & Surveillance System


The Route 8 project is a new 13.5 km section of the Route 8 highway, which connects the existing Tsing Ma Bridge with the town of Sha Tin in the New Territories.

The Tsing Ma Bridge connects Hong Kong Airport on Lantau Island to the Hong Kong mainland. On this new Route 8 section, there are three 3-lane twin tube tunnels, the longest being 2.1km. There is a 1.6km cable-stayed bridge over Rambler Channel, which connects to Hong Kong’s container port facilities, and a small 0.7km 2-lane enclosed highway section.

For providing this complex stretch of highway with the most extensive and advanced TCSS in the region, there were basically two major challenges – the size of the system and the complexity of the traffic operations. The central TCSS system was responsible for managing all of the field subsystems. There were over 3,000 field devices from at least a dozen different vendors and the goal was to provide integrated operations. As such, this represented a massive integration and commissioning exercise. In terms of complexity, the Hong Kong government recognized that it was virtually impossible for a human operator to effectively respond to a traffic situation where potentially several hundred signs and signals were involved. And because of the complexity of the road network and tunnel contraflow operations, there were tens of thousands of possible traffic scenarios, so using a predefined plan approach was not feasible.

We provided a rule-based plan method, whereby a finite set of rules would deal with all possible traffic scenarios. The idea is fairly simple: if dealing with a single stretch of road, a simple rule can be used. For example, for an incident blocking a single lane, close the lane 700m upstream and provide warning to the motorists for the next 1.5 km. But, because the Route 8 network consists of three tunnels all with the capability of contraflow operation plus dozens of interconnecting ramps, there were so many exceptions to the basic rules that it became unmanageable. We eventually came up with a technique able to take into account various operational conditions and combinations with a manageable set of rules. As far as we are aware, the techniques developed have never been used in such a complex traffic network.

 

 

Complexity and innovation

The system incorporates more than 1,000 different types of traffic control signs and signals, such as 15 full matrix variable message signs, 160 variable speed signs, 488 tunnel lane signals, 322 matrix lane signals, and 131 variable prismatic signs. Traffic management, especially for tunnel contraflow and wind management for wind conditions, demands simultaneous control of hundreds of field devices. The intelligent functions in the system greatly reduce operators’ loading and eliminate human errors. To support the complex operations for Route 8, the TCSS has an automatic incident detection function using multiple detection algorithms (APID, McMaster, and Queue Warning) and video imaging technology. The operator is provided with guided event management functions supported by advance rule-based plan generation capability.

The Route 8 TCSS incorporates a rule-based response plan module to respond to incident, queue, planned events (eg, scheduled roadwork) and various adverse weather events.

The TCSS rules for Route 8 reside not only in the decision-making logic in selecting the plans, but also in selecting the appropriate field devices to generate the planned responses. The rule-based response will be generated and recommended to the operator based on the type, severity and location of the event, the operational state of the tunnel (eg, in contraflow) input by the operator or system and the availability of response devices (devices may be ununavailable due to maintenance or equipment failure).
 

Integrated rule-based design

The rule sets in the Route 8 TCSS control all of the visual field response subsystems targeted at the road users, including full variable message signs (FVMS), lane control signals (LCS), variable speed limit signs, prismatic variable message signs, ‘turn on radio’ signs, and radio messages (such as radio break-in messages in tunnels).

To provide greater flexibility to address various objectives of the field devices, customised display rule sets are defined for specific individual type of field devices.

The rule sets for some devices are more complex than others. For FVMS they define the sign selection criteria, the message type, the message format and the display of the appropriate message in a single step. The rule sets for LCS define the signal selection criteria, the display aspects, as well as a safe sequencing of response (both spatial and temporal) to change from the current signal displays to the appropriate display aspects. The sequencing of response of rule sets for LCS is also required to account for various lane configurations and the different lane blockage patterns.

All of the customised display rule sets are designed by the traffic engineers based on a set of operational principles that account for various human factors. Moreover, the rule sets for various field devices can be applied at the same time, and the responses will be consolidated and prioritised based upon other sets of arbitration rules. Special principles are also defined to account for transition areas, such as tunnel approaches and exits.

Many systems only have rules for individual device types, such as lane and speed control. In North America, rule-based design is mostly limited to dynamic message sign operations. The Route 8 rule-based engine employing XML technologies is the most flexible architecture to support integrated operations that seamlessly handle real-time mix of lane management, speed management, warning, passive and active diversion, etc. If an agency is looking for error-free, consistent operations of a complex active traffic management system with various forms of interacting traffic management strategies, this architecture offers flexibility to implement the control measures that they want.

For more information on companies in this article

Related Content

  • In-vehicle vision-based systems and autonomous vehicles
    January 11, 2013
    The Artificial Vision and Intelligent Systems Laboratory (VisLab) of Italy’s Parma University has built itself a fine pedigree in basic and applied research which has developed machine vision algorithms and intelligent systems for the automotive field. In 1998, a VisLab-equipped Lancia Thema named ‘Argo’ travelled along the famous Mille Miglia race route and completed 98 per cent of it autonomously using then-current technology. In 2005, VisLab provided the vision element of the Terramax, a collaborative un
  • The rise and rise of robo-car
    July 23, 2019
    When it comes to driverless cars, there are many variables – but one thing is for certain: autonomous driving will have a significant impact on vehicle design, says Andreas Herrmann The transition to autonomous vehicles (AVs) means that many of the factors which have shaped automotive design for the past 130 years no longer apply. At present, the design of a car is largely determined by the anticipated direction of travel: the car’s silhouette immediately shows where the front and back are. Driverless ve
  • Standardised technology aids low cost wireless communication
    November 13, 2012
    In the UK, the necessary radio spectrum has been identified and standardised technology developed to allow cost effective wireless communication between cars, devices and other ‘machines’. This by Professor William Webb. A world free of traffic congestion, with intelligent systems directing vehicles and alerting drivers to free parking spaces may sound a far off fantasy to motorists stuck in seemingly endless queues on the outskirts of London. Yet this is a scenario not confined to the world of science fict
  • Machine vision needs standards to fulfil ITS demands
    May 28, 2014
    No-one should expect the enabling qualities of machine vision to come free of charge but Jason Barnes finds there is still much that ITS stakeholders can do to help reduce costs. After many years of application in high-end solutions for the enforcement and tolling sectors, machine vision is gaining traction in more general areas of traffic management. Nevertheless, those OEMs producing transport-oriented solutions which incorporate machine vision and looking to increase the technology’s share of the ITS mar