Navtech Radar monitors traffic on Swedish highway

For the surface installation, Navtech installed two TS350X model radar systems.

The Swedish Transport Administration, Trafikverket, has installed the Clearway incident detection system on a 16 km section of the highway E4. This covers all lanes on both carriageways and automatically sends alarms to the control centre when incidents occur.
Trafikverket is responsible for the operation and maintenance of the Swedish road network. The harsh winter climate in Sweden presents the Swedish Transport Administration with particular challenges when it comes to operating and maintaining the road network. Poor visibility, caused by winter rain, snow and fog, is a common cause of vehicle collisions. In addition ice, which is common, causes damage to road surfaces, making driving difficult and leading to higher than expected numbers of slow or stopped vehicles, or even pedestrians. This is true in heavy traffic, where incidents need to be addressed quickly to avoid more general road network disruption. Light traffic also gives rise to problems, as vehicles tend to move at greater speeds and the potential for major accidents, as the result of collisions with stranded vehicles, is increased. Timely detection of incidents and posting of warnings to upstream drivers via motorway message signs is imperative, therefore.
In recent times, traffic operators in the control centre in the Swedish capital have used CCTV to watch for incidents. Alternatively, reports have come via phone, either from those directly involved in an incident or from passing motorists. Clearly, this is not a reliable way to detect all incidents quickly, to maintain lane throughput, or to look to reduce the numbers of incidents and accidents. A solution was sought.

Many requirements
Trafikverket had a number of requirements in addition to reliable detection irrespective of the general weather conditions. It needed to ensure that detection was continually reliable, with little or no routine maintenance. Full lifecycle costs are an important consideration when designing and financing an incident detection system and, as well as equipment failure or malfunction, operating costs should also include the nuisance cost of responding to system-generated false alarms.
Trafikverket also required the system to provide full-carriageway, all-lane coverage from only a few installed sensors, as the costs of installing and providing power and communications to a large number of additional gantries or mounting posts could not be justified. Long-range detection capabilities were essential.

Two trial locations
In early 2010, Trafikverket invited UK company Navtech Radar to demonstrate the capabilities of its Clearway radar incident detection system. Two trial locations in the Stockholm area were chosen, one above ground on the E4 and another below ground in the Södra Länken tunnel.
For the surface installation, Navtech installed two TS350X model radar systems. The TS350X can detect a man-sized object out to a radius of 350 metres. Since it can provide 360 degree coverage, each sensor can therefore provide a total of 700 metres coverage where conditions allow.
To limit costs and reduce works by taking advantage of nearby power and Ethernet connections, the radar systems were each installed on a post close to an existing gantry. In addition, a number of fixed cameras were installed high on the gantries. These were to provide video images of detected incidents for the purposes of testing and evaluation.
The data from the radar was communicated back to the client's office where a processing system was configured to generate incidents alarms including: stopped vehicle; slow-moving vehicle; pedestrian; reversing vehicle; and lost cargo/debris detection. These alarms were subsequently used to trigger recording on a local digital video recorder. After an initial configuration period the system was left to run and detection and false alarm rate performance assessed.

Examples of typical detected incidents were registered during live traffic operations. In each case the radar systems generated the alarm messages and camera pictures were taken from recorded video. In some cases the weather conditions were too poor for a good-quality video image. Nevertheless, the radar systems generated valid alarms.
Typical false alarm rates during the course of the trial were measured and found to be around one for every 24-hour period. It is hoped that these will be further reduced during a planned rollout of the system as processing algorithms are improved and further installation experience is gained.
Since the Clearway processing system is able to track each vehicle as it travels along a road, during the course of the trial Navtech's technical staff were able to adapt the processing to suit the client's requirements. For example, peak-hour traffic queues on the section of the E4, although technically involving stopped vehicles, and therefore incidents, are not of interest to control room operators dealing with emergency response. They are only really of interest for the purposes of general management of the road network and perhaps future planning. As 'incidents' they simply add to the already long list of alarms that operators have to deal with.
Processing business rules were added that allowed the number of vehicles below a threshold speed to be counted. In the event that these exceed a threshold in a defined length of the carriageway, a queue flag can be set (and later unset). These alerts are made up of several stopped or queuing vehicles, as is usually the case in a traffic queue, but do not require an immediate operator response as in an emergency situation. Of course, if a stopped vehicle is detected outside of the queuing lane then the incident will continue to be detected.

Product Suppliers
Back to top