Robustness is the extent to which, under pre-specified circumstances, a network is able to maintain the function for which it was originally designed.
Generally, this function is designed to enable traffic flow from A to B with a certain speed. When a network is robust, this means that even in the case of unexpected events, the network is able to accommodate the traffic flow with a sufficiently high speed. For example, if an incident (or terrorist action) happens that blocks a certain road, there should be sufficient alternatives routes with sufficient spare capacity be available to reroute the traffic over these alternatives. There are certain criteria that an urban planner can take into account in order to design a robust network, or to test if an existing network is robust. Some possible criteria will be discussed later.
Forms of robustness
Robustness can be encountered in two different ways, namely static robustness and dynamic robustness. In traffic, mainly dynamic robustness is of relevance. Dynamic robustness refers to the case in which the dynamics of redistribution of flows should be taken into account. Obviously, the road network is an example of a complex network in which dynamic robustness should be considered. For instance, an accident could block a road completely or partially, which causes delays for the road users, and some of the road users will change their route choice or maybe even their departure time choice, mode choice or destination choice. Robustness of a traffic network can furthermore be distinguished in the following concepts:
The robustness of a system can be increased by introducing a certain spare capacity into the system. This spare capacity is often referred to by the term redundancy. When disruptions occur, this spare capacity can be made available so that the system continues to function better. A traffic network has sufficient redundancy if there are alternative routes available and there is enough spare capacity on all the routes. There are actually two types of redundancy: active and passive redundancy. Active redundancy, like alternative routes, is redundancy in the network that can also be used in regular situations. Passive redundancy refers to back-up options that are only used in case of disturbances. An example of this are ferries that can be used in case of bridge failures.
The robustness of the system can partly be measured by the degree to which the system is able to fulfil more and different functions than the functions for which the system was originally designed. In other words, flexibility is a property that enables the system to expand in line with new requirements that are demanded of the system. A flexible network is able to reroute the transported elements to the alternative routes as fast as possible, for example by using both driving directions of the road in case of a mass evacuation.
Resilience is the capability of the transport system to recover, preferably within a short time period, from a temporary overload. For example, with good traffic and incident management strategies, the emergency services are able to arrive at the incident spot and remove the blockage as soon as possible. Resilience can be divided into two types of system resiliencies:
- static resilience: the capability of a transportation system to maintain its function.
- dynamic resilience: the capability of a transportation system to recover rapidly from a severe shock to achieve a desired state.
A more general description of resilience in the context of societal security is given here.
This is the degree to which traffic congestion remains restricted to the relevant link or a small section of the network. If there are less interdependencies in the network, congestion at a centrally located link or node will not cause a series of cascading failures disrupting traffic on large parts of the networks. With compartmentalization it is possible to reduce spillback effects to a minimum level. For example if there is congestion on a certain road, when it moves downstream until an intersection, the congestion will normally spread to the connecting roads. Compartmentalization should prevent that the congestion spills back unto the next intersection, e.g. by opening an extra lane to buffer the traffic.
The probability of a road network performing its proposed service level adequately for the period of time intended under the operating conditions encountered.
Vulnerability is the opposite of robustness; a vulnerable network is vulnerable to disruptions, and will not be able to accommodate the traffic in case of abnormal traffic situations.
A more general description of vulnerability in the context of societal security is given here.
Method to detect robustness issues in a road network
In order to indicate at which locations in the network measures concerning robustness are potentially the most effective, TNO developed a method to quickly detect vulnerable road sections. The criteria of this method could be used also by the urban planner. It is not yet adapted or focused on security issues.
It assumes a good knowledge of the network and traffic performance of the network under consideration. This can for example be gained by measurements or traffic simulation. If no traffic data are available yet or the network does not yet exist in reality, the urban planner needs to have a good knowledge of traffic processes, or perform simulations in advance.
The method consists of the following 4 criteria:
- Speed ratio: the speed ratio is the ratio between realized speed and maximum speed. As the speed ratio decreases, that part of the network is more vulnerable. A low speed ration indicates congestion and so few spare capacity on the route itself.
- Ratio flow/spare capacity on alternative routes: when alternative routes with sufficient spare capacity are present, the network is more robust. The higher the flow, the more spare capacity is needed to handle disturbances. The closer the ratio is to zero, the less vulnerable the road section is.
- Hindered vehicles from spillback effects: if a disturbance on a certain location quickly leads to large spillback effects, this location is vulnerable.
- Probability on disturbances: as the probability on a disturbance on a certain location is increasing, that location is more vulnerable.
The four indicators together determine the vulnerability of a road section.
Footnotes and references
- (Rose, 2007; Transportation security and the role of resilience - A foundation for operational metrics.pdf)
- Billington and Allan (1992) and Wakabayashi and Iida (1992)