41 
Hood Section 
b ~ 13 m 
Fig. 9. Transverse section of the road embankment. 
The GPS+GLONASS RTK survey consisted of: 
• repositioning the planned sections on the ground, 
during the various repetitions; 
• measuring their variations. 
The measurements were taken twice a week for six 
months. The bi-weekly survey of a road embankment was 
performed in this way. The operator positioned the 
Master receiver on a vertex of known co-ordinates and 
initialised the Rover receiver, with a standing time of no 
more than one minute. At that point, he performed the 
operation of plane calibration: within the GPS reference 
system, he measured the co-ordinates of at least five 
points on the ground which had co-ordinates known also 
in the local system. When the double set of co-ordinates 
was entered into the CDU-2, the parameters of the plane 
calibration for passage from the GPS+GLONASS co 
ordinates to the local ones were calculated (Fig. 10). The 
maximum differences in the calibration were never 
greater than 5 cm. 
Fig. 10. Display of the Field Face Program in the CDU-2. 
When the parameters of the calibration and the planned 
co-ordinates of the sections were stored in the data 
logger, the operator proceeded with the measurement of 
the points in transverse directions. During the 
measurement operations, the acquisition interval of the 
single receivers was set at 1 sec and the minimum cut off 
of the satellites at 5°. The survey of a single road 
embankment, including the operations of repositioning, 
lasted from 1.5 to 2 hours. In this time period, 15 to 20 
sections were measured, i.e. 300 to 500 points. 
The number of visible satellites was never less than 12- 
13: on average 8 GPS and 5 GLONASS. The PDOP was 
never greater than 2. In this way, it was sufficient to stay 
with the Rover receiver on the survey points for no more 
than two epochs of measurements (Fig. 7 and 8). The 
same survey was performed with equipment similar to the 
previous one but with receivers for only GPS 
measurements in double frequency L1-L2 (GPS RTK). 
The sampling interval was the same (1") and the cut off 
was increased from 5° to 10°. With this equipment, the 
interruptions of the signal were more frequent, especially 
in correspondence to the scarps of the road embankments 
which reach considerable heights and inclinations (Fig. 
9). The number of dynamic re-initialisations was 
increased and with them also the respective times of static 
(of initialisation and of a single fix) and dynamic 
stationing. Moreover, to permit the Rover device to 
receive many satellites and therefore to record the 
position, we were often forced to leave the area of 
measurement, with a considerable increase of the 
difficulty of surveying. From the operational point of 
view, the experiment showed: 
• the considerable capacity for acquisition of the 
GPS+GLONASS RTK system, especially in 
conditions particularly unfavourable for reception of 
the signal, e.g. during the survey of scarps or the 
calibration on vertices situated near electric power 
lines; 
• the good reception of the radio link signal from the 
Master to Rover, which never suffered interruptions 
that might compromise the execution of 
repositionings and fixes; 
• the reduced times of static initialisation and dynamic 
re-initialisation; 
• the good planimetric and altimetric precision, always 
within the tolerances established by the client; 
• the limitations of the GPS RTK system, both in the 
phase of initialisation and the phase of acquisition. 
Table 1 shows a comparison of the measuring times and 
number of operators employed in the two satellite 
techniques, as well as in the survey with an integrated 
theodolite, for an equal number of points measured (300- 
500). It can be seen that productivity was increased by 
about 30% with respect to the traditional instrumentation, 
while the presence of operators was reduced to the 
minimum (only 1 GPS or GPS+GLONASS operator 
versus 1 theodolite operator + 2 rod holders). 
Measuring techniques 
Measurement time 
(minutes) 
Operators 
Integrated theodolite 
240-300 
3 
GPS RTK 
120-180 
1 
GPS+GLONASS RTK 
90-120 
1 
Table 1. Comparison of the measurement times and the 
number of operators for the different survey methods. 
2.2 Localising reference points in woodland for a 
water supply tunnel 
The second application involves the checking of the co 
ordinates of some vertices of an axis network for the 
organisation of the course of a water supply tunnel. These 
vertices are situated in an inaccessible wooded area,