132 
The method is very fast (up to 1 point/sec) but applied 
to the survey of road trajectories has one inconvenience 
that is unfortunately very frequent: the loss of the signal 
due to the satellites being concealed by obstacles along 
the road (trees, bridges, houses, ...). This means that 
the system must be re started and notably increases 
the survey time, and costs, but above all the decrease 
in the precision to the metric level, as the interruption of 
the signal induces errors in the successive 
measurements. 
Frequent concealment of the satellites is possible 
above all along local roads, which is the greatest part of 
the system and needs the most urgent functional 
classification. For these roads, which are often lined 
with trees or that cross urbanized zones, the precision 
surveying with GPS kinematic is surely hindered by 
unfavorable road side conditions. In some cases, the 
survey with GPS could not be possible due to the 
difficulty to receive the signal from four satellites at the 
same time for long stretches of road way (built up 
areas, deep trenches, long lines of trees ). 
It is therefore necessary to integrate the survey with 
traditional topographic techniques (theodolite and 
distance measurement) to connect the stretches 
surveyed by GPS. An alternative, especially for 
complex geometry, is to use GPS in conjunction with an 
inertial navigation system (gyroscopes laser 
interferometry). The survey, however, loses its most 
interesting characteristics for the use in the field of road 
networks: speed and precision. 
Table 4 - Methods of differential quick-parking surveying (literature data) 
METHOD 
PARKING 
PRECISION 
CONDITIONS OF USE 
rapid-static 
rapid: 
5 -20 min. 
0.5 cm ± 1ppm 
Receivers with dual frequency needed for rapid- 
static surveying. 
pseudo-static 
rapid: 
5 + 5 min. 
2 cm ±1ppm 
For greater precision it is necessary to return twice 
to the same point. 
Kinematic 
Very rapid: 
1 sec. 
2 cm ± 1ppm 
It is necessary to initialize the system and not lose 
contact with the four satellites for all the session of 
surveying 
There exist dual frequency receivers especially 
designed for real time kinematic surveying, which, 
following an interruption of the signal from the satellites, 
carry out an On-The-Fly {OTF) restart, without the need 
to suspend the survey and without lowering the 
precision. 
The real time OTF kinematic method should make it 
possible to survey in cases of frequent loss of signal 
due to natural or artificial obstacles, but its reliability for 
road networks must still be proven. 
The possibility of defining the plano-altimetric trend of 
the road axes by interpolation or the approximation of a 
number of discrete points to a polynomial B-Spline 3D 
line, suggests that it could be well used, at least in 
certain cases, by methods that can be defined “quick 
parking ”, such as the rapid-static method and the 
pseudo-static method (tab. 3), integrating the survey 
with traditional topographic methods. 
The lower number of total points to survey (according to 
a distribution dependent on the geometry of the stretch 
and the final precision) could compensate for the 
greater time needed at each point. In any case, to the 
greater accuracy of the pseudo-static and rapid-static 
methods respect to the kinematic one, there is the 
possibility of positioning the receiver directly on the road 
axis or its verges, avoiding the passage of the double 
trace of the (presumed) center line of the lanes of the 
carriage-way. 
Whatever the method applied for the definition of the 
geometry of the axes, the survey of the road must 
always be completed with discrete measurements of 
the width of the platform with traditional topographic 
methods (theodolite and distance measurement). 
As regards the altimetric profile, it has been 
experimentally found that the centimetre precision in the 
determination of the planimetrie trajectory of the vehicle 
leads to the same accuracy on the third dimension and 
that, in general, can allow the surveying of the 
superficial macro-deformations. It is useful to point out 
that this cannot be the aim of surveying with GPS, as 
instrument and investigation methods exist exclusively 
for the surveying of superficial characteristics of the 
road surface, which offer results much more precise 
and immediate. 
6. Measurement Experience 
About 40 km of road was measured with a pair of 
Trimble 4000SSÌ, where one was stationary (base 
point), and the other vehicle mounted (rover) referring 
to only one base point, for a duration of about 3 h. 
The measurement was carried out kinematically with an 
acquisition rate of one point per sec at a vehicle speed 
of 40 km/h (total points about 4,000). The rover 
instrument was set-up for a PDOP <7 and Elevation 
Mask = 13. Data was elaborated by GP SURVEY 
software in post-processing. The precision was 
centimetric, some stretches were sampled by 
conventional means as a control. 
An experimental section of about 7 km was measured 
in both directions obtaining two almost parallel curves 
(Fig. 1). We are currently working on an algorithm that 
will allow the definition of the geometry of the road axes 
with centimetric precision, the results will be published 
in the future. 
Instrument limits were: 
■ signal loss in cuts, and in the presence of fixed 
obstacles at the road margins (trees, sustaining 
walls, buildings, etc);