Technical description of the multi-function ARAN equipment subsystems
Measurements of longitudinal road irregularities
(International Index of Road Irregularities – IRI: EU, US parameter-certified by Second International Experiment PIARC – FILTER/EVEN)
Detailed description of the measurement’s principle
The measurement principle used by the ARAN equipment lies in contact-free scanning of the values of vertical acceleration of the measuring axle’s un-sprung mass and the values of vertical acceleration of the sprung mass of the vehicle’s body. ARAN uses state-of-the-art technology for scanning values required to calculate the IRI by a commonly used model known as a quarter-vehicle, defined by internationally-accepted studies (First International – Brazilian Experiment, Technical Regulations of the World Bank No. 46, PIARC International Experiment - FILTER 1998). This method is based on a simulation of a “quarter-vehicle” through computation from the values measured by a laser camera, the so called comparative vertical plane, generated by the optics system of a camera, rigidly mounted on the mass of the vehicle’s body, generating a longitudinal road profile at 50 mm intervals, detecting wavelengths from 100 mm to 100 m, approaching the so called true profile, and from the values obtained from an accelerometer, installed on this camera, measuring gravitational accelerations of the vehicle’s mass in interaction with the road’s longitudinal irregularities. The result is a more accurate IRI value, approaching the exact computation of a profile measured by a classic land surveying method, and in step with the latest trends in this field, presented at the last PIARC International Congress. The readings obtained from the profile and acceleration measurements are pre-processed by the designated subsystem’s control computer and sent to the central board Pentium III computer, which in real time computes the IRI index, which the operator sitting in the vehicle can observe. All values are stored on storage media for further processing in office workstations. Positions of all measurement readings are digitally and visually identified according to the client-defined methodology, both in linear as well as nodal systems. Positioning of measured parameters is described in a separate section – refer to Positioning of all Vehicle’s Subsystems.
ARAN meets all requirements on IRI measurements specified by Clause 8.3 of the ÈSN 73 6175 standard. Our vehicle is capable of measuring in accordance with the requirements specified in this standard, i.e. by a pure response method using two accelerometers. However this method is now regarded as obsolete, and it is recommended that the profile measuring method be used instead to obtain values for the modern algorithms that are used in the US according to Class I of the ASTM standard and applied for the measurement of this parameter in all contracts for the Czech Road and Freeway Board since 1997.
Measurement accuracy
The accuracy of the used accelerometer is ± 1 mG, and that of the 16 kHz laser camera ± 0.2 mm. The accuracy of the resultant IRI measurements with the so-called golden model, computed on test sections in the US from the so-called true profile, is expressed in terms of a correlation coefficient guaranteed by the manufacturer, R2 = 0.965. The same correlation was achieved with this specific equipment in a Czech national experiment.
Measurement speed
Thanks to the new algorithm (Low Speed IRI) and the type of accelerometer used, the measurement speed can be variable, ranging from 25 to 90 km/h. A typical operating speed when measuring roads in the Czech Republic is between 40 and 80 km/h.
Sampling frequency
The sampling step is 50 mm, which allows the system to pick up wavelengths from 100 mm to 100 m. The IRI values are then determined as an average value of the profile, adjustable from 10 m upwards. The vehicle’s measuring track can be set as required. However, the selected position of the measuring track has to be taken into account, because in the left track, i.e. the one closer to the road’s centreline, the IRI values are, particularly in Class I and Class II roads, significantly lower than on the right side of the road, where due to the lower rigidity of the road’s hard surface near the edge, the IRI values will be considerably higher, and hence there is a certain danger that, should this track be changed, the existing development series will be distorted.
Measurements of transversal road irregularities
(depth of wheel ruts, theoretical depth of water in wheel ruts, cross-section gradient – EU, US parameter-certified by Second International Experiment PIARC – FILTER/EVEN)
Technical description
Cross-section measurement are performed in accordance with the requirements of Clause 5 of the ÈSN 73 6175 standard, using ultrasonic sensors, mounted on a 190 cm long, rigid bar in front of the ARAN diagnostic vehicle, and on two pairs of auxiliary mounting adapters, 60 and 90 cm long, respectively. Signals sent by the sensors are continuously processed by the system’s control PC, mounted on the same rigid bar that holds the sensors. The control computer sends processed data to the ARAN diagnostic vehicle’s central computer, where they are stored together with other data on the computer’s hard disk. From the data sent by the sensors, combined with data from other measuring subsystems (gyroscopes and distance measurements), are then in real time computed relevant variable parameters of the road surface, i.e., the depth and type of the wheel ruts, a theoretical depth of water in the wheel ruts and the cross-section gradient of the road, and the resultant values stored in a data file.
Accuracy of transversal irregularity measurements:
Measurement accuracy of each ultrasonic sensor is 1 mm.
Measurement accuracy of gyroscopes is 0.1%.
Measurement accuracy of electronic length measuring device is 0.1%.
Speed of transversal irregularities measurements:
Transversal road irregularities can be measured at speeds between 0 and 90 km/h.
Frequency of transversal irregularity measurement sampling:
In transversal direction, the frequency of sampling is determined by the location of sensors on the main bar or on the auxiliary adapters, which is 10 cm. This distance between the sensors in the transversal direction seems to be optimal in order to obtain reliable measurements of wheel ruts, and to identify their type. The width of the measured transversal profile is selectable from 190 cm, 250 cm, and then in 30 cm increments up to 370 cm, depending on the width of the driving lane’s hard surface section (Selecting an optimal width of the actual measurements is also necessary for accurate measurements of the road’s cross-section gradient). Sampling intervals in the longitudinal direction are also selectable, starting from 5 m in 1 m increments, as required by the client. (The typical interval is 10 m).
Limitations of the transversal irregularities measuring system:
From the physical principle of the measuring with ultrasonic sensors and from their design, it is obvious that these sensors cannot be used in rain on a wet road, and also dirt on the road can distort the results. It was established by experiments that when measuring road surfaces with a negative texture, about a 3 mm correction has to be applied to the wheel rut depth. The shortest sampling interval is 1 m.
Measurements of road surface macro-texture
(ISO MPD – EU, US parameter-certified by First International PIARC Experiment)
Detailed description of measurement methods (principle)
The principle used by the ARAN measurements system is based on regular contactless scanning of longitudinal profile samples, from which the system computes an average depth of the macro-texture, MPD (ISO-certified). Longitudinal sections are scanned again by a laser camera made by SELCOM of Sweden (practically the only manufacturer whose products are used in almost all similar devices).
The system is mounted on a vehicle carrier above the right or the left wheel track where the values are critical, measuring by certified high-speed lasers an average depth of the MPD macro-texture and a macro-texture profile of periodical 20 cm long road surface samples. Texture data are an important component of the road’s anti-skidding properties, and are used to decide whether other measuring systems should be deployed, such as skidders. Correlation of this parameter was done during the First International Experiment in Belgium and Spain and, similarly as in our national experiment that had taken place before the parameter was included in the Ostrava road databank, shows in the ARAN system a very good correlation with a macro-texture test carried by the sand patch method, namely R 2= 0.98.
Profile measurement readings are pre-processes by the control computer and sent to the central board computer, where the algorithms of the MPD and RMS values (mean quadratic deviation) are computed in real time, and which can be monitored by the operator in the vehicle. All values are stored on storage media for further processing on office workstations, and are used for further processing of detailed computation of texture profiles or homogenised statistical summaries for expert applications in road management systems. Both linear and nodal positions of the measurement reading points are identified digitally and visually (electronic header in the video recording) in compliance with a client-specified methodology. The methods of positioning the measured parameters are described in a separate chapter, Determining the Position of All Vehicles Subsystems.
A number of comprehensive studies and tests (First PIARC International Experiment, Czech National Experiment and other) confirmed the correlation of the MPD parameter with the standard ASTM sand patch test as being 98%, which was the reason for including this parameter into the international coefficients IFI and EFI, and also into the ÈSN 73 6177 standard.
ARAN meets all requirements for MTD measurements as specified in Clause 6.3 of ÈSN 73 6177 (as suggested by the chapter’s title). Our equipment is also capable of measuring the MPD parameter by the newly certified ISO methods.
Measurement accuracy
The texture-measuring module comprises one or two 62.5 kHz laser camera pinpointers, working with an accuracy of ±0.1 mm.
Measuring speed
The system works for all wave widths of the macro-texture at speeds up to 90 km/h.
Sampling frequency
The resultant MTD and RMS values are stored at intervals selected by the operator, typically 20 m at the profile length of 0.2 m. The profile samples are scanned in a laser camera frequency (in ARAN, the frequency of measurements – the distance between individual surface reading points – is 0.45 mm). For other applications, detailed surface profiles are stored too. The measuring vehicle track can be set as required, i.e., in the left or the right vehicle wheel track, where the macro-texture values are understandably worst, and have the highest effect on the vehicle’s interaction with the road surface. In our case it is in the right hand side, as on the left hand side is mounted the IRI system. Nevertheless, the systems can be located anywhere, as required by the client, although definitely not above the middle of the driving lane, where the macro-texture values are better than in the wheel tracks.
Determining the Position of All Vehicles Subsystems
Detailed description of the measuring method (principle)
ARAN uses a unique three-element positioning system that has been adopted worldwide by all manufacturers of comparable equipment. The main element is based on the principle of electronic scanning of the driven distance by a very accurate, calibrated Henstler sensor, located on a vehicle’s wheel, and linked to the central computer that writes the distance covered from the starting point (node) in 1 m increments into each file of the measured parameters, thus unequivocally linking all measurement readings to the database oriented by this way.
The second element is the video recording itself, marked with an electronic header containing the above-described stationing data, plus the name of the starting and end point of the measured route, and other identification data such as an instantaneous vehicle speed, direction, driving lane, road number, etc. Part of the major data in this header is also the number of a frame of the front video camera that is, as the distance data, assigned by the header file to each measured parameter, and that serves as a positioning element for other system modules, such as for instance the Visidata database viewer, linked to the digitised video recording of the given locality; and by means of the last positioning element, DGPS coordinates, also to the map, in which the measuring equipment’s instantaneous position is graphically represented. Thanks to the interconnection of all three positioning elements, this viewer is very user-friendly and simple to use. This module is used by a number of administrative institutions in the US, where collecting this type of data is very common.
For this purpose ARAN is equipped with a POS/LV system made by APPLANIX of Canada, which is integrated with the ARAN system both hardware- and software-wise, and which allows, by three high precision optical gyroscopes in conjunction with three precision accelerometers and a pair of satellite receivers working in the GPS differential mode, the supply of very accurate information about the vehicle’s inclination is all three axes (useful for obtaining more accurate transversal and longitudinal inclination parameters); and after processing satellite signals and computations by means of relatively complex algorithms based on the Kalman’s filters, determination of the instantaneous position of the vehicle in cartographic coordinates with the accuracy of 1 – 5 m. This positioning method is as yet not required in the Czech Republic, and with the existence of the current nodal positioning system is superfluous.
Measurement accuracy
The basic precision, determined by the limitations of the process of calibration that is carried out at a stretch of a known length prior to each departure, is better than 1 m per kilometre. The accuracy of other positioning elements is identical or slightly worse (DGPS).
Measuring speed
Measuring speed is identical as for the other subsystems, although in this case there is no theoretical limitation (again with the exception of GPS, where the rule the lower the speed the better applies, which is unfortunately the opposite to the gyroscope, where the rule the faster the speed the better applies).
Sampling frequency
The lowest sampling frequency is 1 m.
Measuring and evaluating road damages (Video logging)
Technical description
The system comprises of a front SONY DVCAM digital video camera, recording the road surface from the driver viewpoint, and a rear SONY DVCAM digital video camera, recording the road surface from a vertical angle. The rear camera is equipped with two synchronised stroboscopic discharge lamps that guarantee a standard light quality even during variable light conditions.
The video recording includes a sound track recording driver’s and operator’s comments describing the state of the road surface. Video signals of the two cameras are mutually synchronised by means of a time code, which enables subsequent assessment of the road damage on a PC. The combined video recording has also a header added, containing basic information about the measurements (measured route identification, stationing, measuring speed, video cassette number, etc.). Simultaneously, a pilot file is generated on the control computer’s disk drive, with the frame numbers and corresponding chainage.
Workstation for assessing road surface distresses
The workstation is a combination of an S-VHS studio video recorder and a digital DVCAM video recorder, connected to a Pentium III computer with a 23” video monitor and a special keyboard, the keys of which can be individually programmed to evaluate the road surface distress as specified by the client, e.g., according to TP 82. Experienced operators (minimum 5 years practice in damage assessment) assess on these workstations the damage at a certain section, converting the file into a desirable data format.
The original recordings from both cameras and the electronic header are recorded in the vehicle by an S-VHS video recorder for damage assessment purposes, equipped with appropriate viewing devices for clients, and in parallel onto a DVCAM digital video recorder for digital archiving which enables the material’s further use as described below:
In order to satisfy the requirements for improving user comfort and to make economic use of the data contained in the “road passport” database, and upon request received from specialised departments of the Czech Roads and Freeways Board, we have created this subsystem of our own.
The existing S-VHS video recording equipment in the ARAN system was this year complemented by a parallel DVCAM professional digital recording system. The system guarantees a high quality sharp picture and its digital transfer to standard CD ROM computer medium with compression characteristic 1:1, i.e., with a 0% distortion. The existing S-VHS recording system, for which several laboratories have viewing equipment, will be of course kept, but thanks to the new digital DVCAM colour cameras (in ARAN mounted in front and at the back), we will be able offer a final picture of a considerably better quality. The digital recording that will be made also using the new DVCAM digital video recorder also allows the making of copies of these recordings without any distortion, i.e., identical to the original (master) copy in all formats as specified by the user, from the lowest 8 mm THS quality up to the semi-professional S-VHS and Hi8. In this case quality means the number of TV lines of the selected system. The highest quality is the offered digital recording on CD media.
Recording system:
Technical description of delivered results:
 The client receives either a CD or a DVD (starting autumn 2001) with a 15-minute continuous video recording, covering a route of between 10 and 15 km. The DVD technology allows the making of a continuous recording of up to 1 1 hours, or approximately a 70 km route, requiring 120 DVDs instead of a 1,000 CDs. So far, individual recordings of 50 or 100 km stretches seem to be optimal for the needs of the road database which, while maintaining the high quality described above, requires ca 150 to 200 CDs per 10,000 km. The video recordings are made in a standard ARAN format used for collecting variable parameters, i.e., a halved picture, in the top half taken from the driver’s viewpoint, and in the bottom half a detailed orthogonal view of the road surface. Thanks to the new DVCAM cameras’ special wide-angle lens, a wider view is achieved from the driver’s viewpoint, plus an unmatched depth of focus intrinsic to the digital technology, sensitivity to light and compensation of disturbing factors. In case of the rear camera that is now also a colour unit and that is equipped with automatic digital control of counter-light compensation (shadow versus lit surface), the picture is sharper, enabling cracks smaller than 3 mm to be detected, with a perfect detection of corrosion, without using stroboscopic lights that are unpleasant for other drivers.
We offer this service as a significant qualitative enhancement of the user value of our products in the area of comfort and economy of obtaining immediate optical positioning of the surveyed locality or the road stretch, as required by the Czech Roads and Freeways Board. The high costs of often long transfers of vehicles required to achieve the same results fall aside, achieving with this product the same final effect from behind a desk with an ordinary computer on top and, with a network, from more than one workstation at the same time.
This system is by its digital principle predestined for other applications too, notably the ones useful in helping top managements to make quick and effective decisions, by making available to them data, both variable and non-variable parameters that can be in the form of tables or graphs inserted directly into video images. This product can also be used for a follow up passportisation of buildings, the state of vertical and horizontal signs and safety elements. These images can be complemented by measurement scales to make measuring off distances from the images easier, and allowing the road alignment to be evaluated, notably visibility ranges.
March 2001
Ing. Václav Bolina
|
