Assessment of High-Speed Weigh-In-Motion Systems by Road Testing for direct Enforcement

Abstract

The ongoing National research project on direct enforcement by WIM, supported by the French Ministry of
Transports (DGITM) and carried out by IFSTTAR and Cerema, aims to demonstrate the feasibility of using
adapted and certified existing WIM systems for direct overload enforcement. This project involves the
participation of three volunteer WIM vendors. Trials have been carried out in lab and on the large pavement
testing facility of IFSTTAR in Nantes, followed by a on road long period test. Sorting algorithms will be
developed to identify the vehicles weighed within the required legal and metrological tolerances. The road test
intends to assess the accuracy and performances of the proposed WIM systems and to demonstrate the feasibility
of certifying and using them for direct enforcement. This paper reports the organization and results of this test
conducted on a concessionary motorway in eastern France. Very promising results are presented which pave the
way to the implementation of high speed WIM for direct enforcement.
More than 20 days of measurements were carried out from October 2015 until autumn 2017. Rented test trucks
as well as the trucks from the traffic flow are used for this test. The rented test trucks are weighed carefully on
static scales and then make several repeated runs on the WIM systems at various speeds and lateral locations in
the traffic lane. Some trucks of the traffic flow are selected by the police, stopped for static weighing during
check periods, and the data are used to assess the accuracy of the WIM systems.
Three WIM manufacturers provided full systems (sensors and electronics): Fareco (France), Kapsch TC
(Austria) and Sterela (France). Fareco installed piezo-ceramic strip sensors, while Kapsch and Sterela installed
piezo-quartz strip sensors. In addition, IFSTTAR-Cerema also installed additional piezo-quartz sensors. All
together, 15 piezo-electric WIM sensor lines are installed on the test site. 11 lines are made with piezo-quartz
sensors provided by Kistler, which may be used as a multiple sensor array or several sub-arrays. The test plan
and the accuracy assessment methodology are based on the European specifications of WIM COST323. The data
analysis is also carried out according to the OIML R134-1 recommendation, using the maximum permissible
errors (mpe). An important task is to identify the measures which fall within the tolerances accepted for direct
enforcement, i.e. ±5% for the gross vehicle weight, and above all for the fully loaded and overloaded trucks.
Other measurements out of the tolerances may be eliminated based on criteria found by the systems.
More than 70 static weightings are carried out every day of testing, which are used to assess the WIM data
provided by the 3 vendor’s systems. The first five measurement days in autumn 2015 were used to calibrate the
systems. Then 15 days of regular measurements were carried out in 2016 and 2017. More than 1000 weights
were measured during this period, showing promising results. The mean relative errors on gross weight are under
±3% for the two best systems (Kapsch and Sterela).