Evaluation of the Benefits of Active Traffic Management Schemes Using Microsimulation Programming

Evaluation of the Benefits of Active Traffic Management Schemes Using Microsimulation Programming


M Millard, Mott MacDonald, UK; P Unwin, Highways Agency, UK


ATM schemes are used to ease congestion in Britain by the use of variable speed limits and hard-shoulder running. This paper explores the development of a Microsimulation model and program to simulate ATM, and successes and further uses of this tool.


A key problem on British roads is the high level of traffic demand, which promises a steady increase into the future. A cost effective, intelligent and efficient way to ease the problems that this causes is to implement Active Traffic Management (ATM) onto the strategic highway network. ATM works by reacting to traffic conditions on the highway and changing driver behaviour according to these conditions. This involves increasing the capacity of roads during peak periods by using variable speed limits and hard-shoulder running, similar that undertaken on the M42 in the West Midlands..

The variable speed limits are triggered by a combination of flow and speed measurements at specific loop sites along the carriageway, which are processed by the MIDAS congestion control algorithm to provide suitable signals on gantries and signs in the surrounding area. The hard-shoulder is used as a live lane when the speed limit drops to a specified value. The status of the hard-shoulder is indicated on gantry signals allowing drivers to use the extra capacity and ease congestion on other lanes.

ATM congestion control has been successfully implemented on the M25 and M42, and is subsequently being rolled-out around the Birmingham Box.

This paper provides an insight into the design and construction of a Microsimulation model to simulate the implementation of ATM on the M6 J4-J5. The study uses a ?base model? constructed in VISSIM and controlled by a Visual Basic for Applications (VBA) program. The VBA program extracts speed and flow data from the model every minute during the simulation run. It runs this data through an algorithm to simulate MIDAS congestion control and then feeds back the results of this into the model.

The inputs back into the model from the algorithm are speed limits at signal locations and the status of hard-shoulder running in the vicinity of the signal. These inputs affect the model run by opening/closing the hard-shoulder, changing vehicle speeds and also providing a visual representation of the signal display on the road.

There are a number of challenges when constructing the base model and these are outlined below:

The base model needs to replicate observed driving conditions and correct driver behaviour for an ATM motorway. This includes merging / diverging directly to / from the hard-shoulder when this is a live lane. It outlines the logic and assumptions behind the VBA program controlling the model, explains the observed data (flow, speed and journey time) used in the base model and the need for extra observations of speed data for the simulation of ATM.

The purpose of the model is to highlight the congestion and travel time benefits of applying ATM and to pinpoint locations where these benefits are most evident. The model acts as a visualisation tool for observing the traffic effects of Variable Speed Limits and Hard-Shoulder Running. Further uses of the model include testing the impact of changes in the design of the ATM measures such as revising the locations of loops and VMS, the length of weaving sections, the use of through-junction hard-shoulder running, the positioning of Emergency Refuge Areas and the flow and speed thresholds at which ATM is triggered.


Association for European Transport