HARMONY IN PRACTICE


A successful pilot on the D5 highway in the Czech Republic has brought very positive results. So positive, in fact, that the mobile management system for dynamic workzone traffic control was then evaluated by the Road and Motorway Directorate who pushed through its deployment on the main Czech highway network, as Marek Ščerba reports

The Czech transport research centre (CDV), with the cooperation of ELTODO a.s., has developed a robust ITS system for traffic control in highway workzones – often scene of serious injuries and road deaths. Five years of development and testing have resulted in a complementary ITS system based on the synergy of several detection technologies, rich data fusion and the VMS network. After much testing of the system functionalities in real life conditions the development work has resulted in the formulation of technical conditions for public procurement in the Czech Republic. Its positive impact on traffic flow and drivers’ change of behaviour was tested in a pilot on the Czech highway D5 (near Pilsen) from September to November 2016.

The pilot covered 2.3km of highway before the traffic closure, in the direction of Prague where the concrete surface layer had been constructed. The piloted system worked 24/7 and all the installed components were in accordance with the Czech technical regulations on traffic equipment.

The system consisted of mobile LED variable message signs, warning and information LED signs and detection data actors (traffic detectors, travel time detectors and FCD data source). The traffic control was performed by providing speed reduction information through the VMS network based on real-time data from detection actors evaluating the actual traffic flow intensity and speed. The VMS were placed at specific “profile” points in the road (seven in total), with each point consisting of two mobile LED signs on both sides of the road.

The key requirements of the system functionality was a remote control by a central SW that uses VMS pictograms based on automatic algorithms working with real-time traffic data, i.e. speed, traffic intensity and the lanes occupancy.

The system worked with set-up events in the following regimes:
– Automatic regime to display signs for congestion detection, traffic flow harmonization and travel time detection.
– Manual control enabling to display the defined sign on the VMS with confirmation from the centre

The basic functions of the system were: drivers’ warning on congestion, traffic flow harmonization (reducing speed before the congestion occurrence) and informing on actual delays. The detectors used in seven profiles (the seventh profile is not designated in figure 1 as its function was just a verification, not a display function) were microwave (Wavetronix), Bluetooth and camera, the communication was made via 2G-3.5G (GPRS – HSDPA) and WiFi.

Another integral part of the system was the power management. Mobile VMS and detection stations at most of the profile points were independent of a direct power supply thanks to a methanol unit power generator (featuring a 27-litre barrel for a five-week pilot power supply) or with accumulators (shifted every four days which has shown as the weakness in this type of energy supply). With two barrels of methanol an ITS station is energy-independent for 10-12 weeks. The bidirectional communication of the roadside ITS stations with the central server was made via GSM/GPRS network. Despite the network availability analysis at every station sometimes the migration from 3G to 2G caused an outage of the system for an average of three minutes. It had no impact on the functions of the congestion warning and traffic flow harmonization but this resulted in a recommendation for a redundant means of communication network for the follow-up deployment. It was also recommended that every control unit should be equipped with DualSIM or with two modems for data transfer. A potential back-up technology could well be CDMA or, where possible, the highway fibre network. Nevertheless, the single SIM technology proved the system reliability and assured continuity of above 97 per cent.

HW and SW evaluation outputs
There was just one failure, specifically a failure of a control unit modem in a profile 2 station, just at the beginning of the test and with no impact on the system performance. Bluetooth detection reliability and the benefits for the system algorithms were evaluated as the Bluetooth detection had been installed on every station and as such provided traffic flow speed at profiles as well as sections. Mobile detection performance has been proven against ASIM reference detectors. The nationwide RODOS system with Floating Car Data (FCD) has been found to be a very valuable back-up system in instances of profile detector outage. It also served for traffic situation reporting and was integrated into the control algorithms of the system as a valuable data source of the overall data fusion.

The reliability of all HW as well as SW components has been proved. GPRS has proven insufficiency to enable it to fulfil the requirement of sending data every minute to the central processing server. The maintenance of the components has also been found to be very easy. One of the main achievements of the system is that the system is independent of VMS suppliers as the key component, the control unit, is compatible with all the detection technologies: it is equipped with a Bluetooth detector and is able to control the connected VMS. It also enables communication with the central server. A unique element of the system is the low power consumption, a maximum of 3W.

Two-phase data collection
There were two six-week phases of data collection, without the system operation (no displayed signs) and in full operation. Based on the Ministry of Transport’s accident statistics (www.jdvm.cz) there were seven accidents without system operation and two accidents with the system in operation (although it should be noted that the two accidents happened on the last day of the pilot). The benefits of the system on improved traffic flows were derived from travel time evaluation (from the FCD RODOS system).

The common travel time in the section (value 0) represented the velocity 120 km/h in one minute of travel. If the delay in the segment was 30 seconds, for example, the vehicle passed the section in 90 seconds at an average speed of 80 km/h. The following table summarizes the results and showcases the benefits of the system on improved traffic flow: the left column contains values just of vehicle detection (phase 1), the middle values with congestion warning signage and the right column with values of traffic harmonization signage (phase 2). The results in green shows that congestion signage provides benefits of almost halving the time of delay (less of 25.09 seconds from a total 56.28) and traffic harmonization provided another 4.56 seconds of improvement.

Pilot system testing had a big impact on delay values before the workzone (delay decreased by 52.68 per cent) and thanks to traffic flow harmonization the vehicles were driven more fluently (the delay decreased by 36.86 per cent). Besides delays, speed and driver behaviour had also been evaluated. The following graph illustrates that after traffic flow harmonization was initiated there was a drop in average speed by 9.24 per cent underlining that the data from the first sign/station showed almost no driver reaction on speed reduction.

The system benefits are obvious at first glance: to evaluate the system the total average values were compared. The average traffic flow harmonization system provided a 10.7 per cent congestion reduction. In absolute values this meant an almost 50 per cent decrease in congestion occurrence. Mainly in instances of sudden high traffic intensities, the congestion levels reached 1500 vehicles per hour without the system operation and 1900 vehicles per hour in the traffic flow harmonization regime. Also the occupancy numbers showed the difference. When the system was in operation there was an increase of 2.3 per cent in lane occupancy – in absolute values this represents an increase of the traffic flow through the bottleneck of 20 per cent.

“We have appreciated the detailed data outputs and the analysis that showcased that such a system can bring high benefits for safety as well as traffic flow, and decided to deploy such a system on various sections of the Czech highway network since 2017”, said Jan Kroupa, general manager of Czech Road Directorate.

The key conclusions
All the ITS control components shall have the reliability higher than 98 per cent. The installation of the system should be deployed inside two days at least. The data sources redundancy proved its value; FCD data integration is very efficient for the system and saves the purchase cost of the system. After harmonization schemes were applied the traffic flow speed has decreased by 8 per cent and congestion occurrence decreased by 50 per cent. The same (50 per cent) is valid for the reduction in delays/travel time in the section 1.5 km before the workzone and 30 per cent at the ‘zipping’ place. The traffic flow intensities increased up to 26 per cent in rush hours with the occupancy of 20 per cent higher than without the system. The system helps to handle sudden increases of traffic intensities to keep traffic flowing in the bottleneck. The economic benefits made according to HDM4 methodology counts for €20,000 on the most congested road network sections (40,000 vehicles/day) and up to €10,000 on less congested roads.

Marek Ščerba is ITS systems architect at the Czech Transport Research Centre (CDV) in Brno, Czech Republic

Author