Digital Road Centre Line (DRCL) Survey of the Mackay District

In June 2004, Dynamic Satellite Surveys Pty Ltd (DSS) was commissioned by the Department of Main Roads, Mackay District (DMR) to perform Digital Road Centre Line (DRCL) survey on the following roads:

Road # Road Name Length (km)
10F Bruce Highway 0.516
33A Peak Downs Highway 7.817
82A Suttor Developmental Road 121.919
85C Fitzroy Developmental Road 1.344
512 Marlborough - Sarina Road 189.373
514 Oxford Downs - Sarina Road 27.557
516 Homebush Road 12.581
517 Sarina - Homebush Road 26.276
518 Eton - Homebush Road 10.429
530 Mackay Bypass Road 4.198
531 Rockleigh - North Mackay Road 3.384
851 Proserpine - Shute Harbour Road 33.461
852 Hay Point Road 13.972
854 Mt Ossa - Seaforth Road 15.476
855 Yakapari - Seaforth Road 23.250
856 Mackay - Bucasia Road 4.434
857 Mackay - Slade Point Road 11.824
5124 St Lawrence - Croydon Road 44.335
5126 Koumala - Bolingbroke Road 23.636
5127 Blue Mountain Road 28.351
5302 Maraju - Yakapari Road 24.388
5307 Collinsville - Elphinstone Road 37.189
5382 Crystal Brook Road 24.353
8501 Gregory - Cannon Valley Road 14.114
8506 Mackay - Habana Road 9.908
8509 Sarina - Coast Road 13.028
8554 Yakapari - Habana Road 11.265
8565 Eimeo Road 4.602
TOTAL 742.980 kms

The DRCL survey was captured in such a way that the information produced could be used in a Digital Road Network (DRN) at a later date.

METHOD OF SURVEY
The data capture was performed using GPS Real Time Kinematic (RTK) surveying techniques. This method enabled both position and elevation coordinates to be acquired in real-time and on the appropriate datum by utilising phase data received from US Navy NAVSTAR Satellites.

The data was captured from the receiver, transformed, displayed and managed by the DSS in-house software package NavLog, which also handled quality control and post-processing.

Real Time Kinematic Method
The RTK observation method involved one receiver (base) being set up over a known control point. A second receiver (rover) was mounted in the survey vehicle. Both stations had a radio link. The base station re-broadcasted the satellite information it received as corrections from its known position to the rover. The rover processed this information combined with the information it received directly from the satellites to resolve the ambiguity and obtain an accurate position relative to the base receiver. This method minimised the effects of ionospheric and atmospheric delays, satellite or receiver clock errors and multipath.

Using the NovAtel OEM4 series dual frequency receivers, accuracies of better than +/- 0.05 in position and elevation were achievable in real time. This was dependent on base line length and satellite geometry. Base line lengths throughout the survey were kept to under 10 kilometres to realise these objectives and satellite geometry taken into consideration in difficult areas.

A typical road run involved initially setting up the base at a known location, entering the base station coordinates into the receiver, and activating the base radio to broadcast the corrections.

After a brief initialisation period, the survey vehicle was driven along the centre line of the route at a maximum speed of 30km/h, logging data in real time each second. This resulted in a position being recorded every 8 metres at that speed but typically at shorter intervals due to the slow speed of recording road furniture. This data was written to a raw observation file for backup and checking purposes. When the distance between consecutive points exceeded a user-defined distance (set at 10 metres for this project), and if the quality of the observation was acceptable, the information was written to a final observation file which eliminated redundant measurements due to stoppages, slow speeds or bad quality.

All road features were recorded using a pulse generator. When the pulse was activated the exact position of the GPS receiver was recorded in the observation files along with a time stamp and pulse number. The feature description was then recorded to a voice-activated tape along with the pulse number, which could then be correlated with the GPS position at a later stage.

During a road run, real-time interactive mapping and longitudinal section displays were available on-screen. This proved an invaluable tool to detect and rectify any “spikes” in the profile which occasionally occurred in bad data areas. It also proved useful for navigating to control points, and resuming position after being “off-road.” The on-screen display also featured many tools for monitoring quality and controlling the operation of the GPS receiver.

If the quality of the data fell below an acceptable standard, an audible alarm would activate and red alarm would appear on the screen. Data could not be written to the final observation file until the quality returned. This usually involved stopping and waiting until an acceptable solution was obtained.

Any relevant survey marks along the route were identified using a MapInfo database and the positions entered into the logging program, which calculated and displayed the bearing and distance to the point as a waypoint. When a check point was reached the GPS antenna was removed from the truck and placed on a tripod set up over the mark. The position was logged and written to a checks file.

On long rural roads any intersecting roads were located concurrently with the survey and recorded as “stubs” using a “mark position” key on the screen. Pull down menus were used to quickly store the road name, material, method and orientation for each point on the stub. Stubs could be recorded as a series of points (activated by the user), or as a continuous one second stream of points, which proved useful for such features as roundabouts. In denser urban areas road stubs were recorded as a separate operation at the completion of the main run.

After a distance not exceeding 10 kilometres, or when the radio signal from the base station became weak, a forward base was established by observation from the rover. Once the new base coordinates were established, the escort vehicle returned to the original base to move it forward, and the survey continued. Sections were overlapped for quality control.

Rapid Elevation Meter (REM) and Electronic Compass
The on-board software incorporated a link to a Rapid Elevation Meter and Electronic Compass which were connected via the USB ports. This allowed the operator to switch directly to these instruments when the elevation data became unreliable in wooded or obstructed areas. The raw elevation and compass readings could be written to REM and compass observation files synchronised with an accurate time stamp from the GPS receiver for subsequent post-processing and correlation with existing positional information recorded at the extremities of the bad data area.
The compass was calibrated initially via the software.

This system was tested prior to mobilisation and proved successful, however it was not required during the survey due to the success of the dual frequency receivers (and a little patience) in bad areas.

Post - Processing
As the data was captured in real time, no post-processing was involved to convert the coordinates to MGA94 and AHD

The output from a road run consists of (a) digital voice recordings and (b) Navlog output comprising final observation files (UTM files), Checks files and Stubs files.

The digital voice recordings are transcribed into a spreadsheet recording pulse number, feature code, side, ID (Main Roads), and comments for each feature along the route.

The information contained in the final observation file as recorded in the field contains the following information for each record: GPS time stamp, easting, northing, elevation, quality, standard deviations, distance tolerance, antenna height, n-value (geoidal separation) and pulse number. At this stage the records have been filtered for bad quality and intervals less than the distance tolerance (usually 10 metres).

The data must be further filtered to eliminate points accidentally logged while the vehicle was off the centre line of the road, reversals, double or accidental pulses and any bad positions or heights that have inadvertently slipped through. Reversals occur when the vehicle is backed up to record a missed feature, and double pulses occur when the operator clicks too hard on the pulse trigger, which was found to be quite sensitive.

Using the Process menu in Navlog the file is first converted to the NEW format as defined by DMR. This imposes a centre line chainage to the data based on a nominated starting value (usually zero) and increasing or decreasing chainage. The chainages are calculated by summing the distances between each consecutive point and correcting them inversely for scale and sea level factor so that a true ground distance at that height is obtained. (MGA94 coordinates are grid coordinates that have been projected on to a plane).

The NEW file is then run through a check program that lists each azimuth and distance, visually flagging excessive bends, intervals, qualities and height differences. This listing is synchronized with two plot displays, one showing a map view and the other a profile view as the data is scrolled. The plots may be zoomed to identify specific areas, and various annotations displayed. Any anomalies are readily identified for investigation. Several (iterative) runs are usually required to produce a final clean run.

Once a final NEW file has been obtained, a features file is automatically created which contains only the pulsed features. This in turn is converted to the TXT format for features as specified by DMR. This is a comma-separated text file, which at this stage has only the positional information of the features. The spreadsheet information is then combined with the TXT file to produce a final TXT (or features) file. This file then exists in both spreadsheet and comma separated format.

To handle the intersecting roads (or stubs) the final features spreadsheet is sorted according to feature code and all intersection coordinates are extracted and appended to the raw stubs file. This is to ensure that road centre line coordinates correspond to the terminal points of intersecting roads. The raw stubs file is then run through a viewer in NavLog, where stub information can be easily checked for clarity and accuracy. Road names are also displayed. Once the information is satisfactory, the file is converted via the software to the final comma-separated TXT format as nominated by DMR.

Click here to view the final surveyed network