Real-Time Applications of GPS
by Paul Manson, Trimble Navigation Europe Limited, UK.
| Abstract Of The Paper & The Profile of The Speaker | Speaker Index | Paper Title Index |
Section 1 Introduction
This paper discusses several specific application
areas of real-time differential GPS technology. The paper is
structured as follows:
_ Section 2 discusses GPS accuracy issues, and looks at why
it is necessary to use differential techniques to achieve
high accuracy with GPS This section also outlines the
difference between code-phase and carrier-phase GPS, and
between the application of differential techniques in real-
time and by post-processing.
_ Section 3 provides an overview of application areas for
differential GPS.
_ Sections 4 through 6 examine in more detail the
application areas of:
A) Mapping and GIS Data Capture
B) Vehicle Tracking
C) Mining & Construction
For each of these three application areas, recent
technological advances are discussed.
Section 2 GPS Accuracy & Differential
GPS
There are a number of sources of error which can
impact the performance of a GPS receiver. Some error sources
are system-wide, while others apply only to a specific
operating environment or a specific GPS receiver.
Most systematic errors can be eliminated using a technique
known as Differential Correction, but errors due to
environmental factors or receiver design cannot be
eliminated using differential correction.
System-Wide GPS Error Sources
GPS error sources which are systematic, and which
can be partially or wholly removed by differential
correction, are summarised in the following chart:
As can be seen, the major systematic error source is
S/A, or Selective Availability, which is a programme
administered by the United States Department of Defense, to
deny availability of high accuracy GPS to civilian users of
the system. A 1996 U.S. Presidential directive announced
that Selective Availability will be disabled by the turn of
the century, enabling civilian users of GPS to enjoy
autonomous accuracy of 12-15m. However, even with the
removal ofS/A, accuracies ofbetter than 12-15m will still be
achievable only using differential correction
techniques.
Local GPS Error Sources
There are a variety of GPS error sources which are
local to a particular operating environment, or specific to
a particular GPS receiver design. Environmental error
sources include:
_ Multipath, which occurs when a GPS signal travels through
two separate paths before reaching a GPS antenna on the
ground. In this case, the reflected signal arrives at the
antenna later than the direct signal, and unless the
receiver architecture can eliminate the reflected signal the
receiver will compute an erroneous satellite
pseudo-range measurement, leading to an inaccurate GPS
position.
_ Satellite geometry (PDOP). When satellites are spread
evenly across the sky, a set of pseudo-range measurements to
these satellites has a good geometry for trilateration, the
mathematical operation of computing a position on the ground
given the position of the satellites and the pseudo-range
distances to these satellites. When satellites are close
together in the sky, the trilateration geometry is not so
good, and measurement errors tend to compound, leading to a
poor computed GPS position.
Receiver-related error factors can include:
_ Receiver channel noise
_ Receiver clock errors
The best GPS receivers are internally 'clean', with respect
to radio-frequency interference, and are also resistant to
external RF interference or jamming.
Local error sources cannot be eliminated by
differential correction, and can only be resolved using good
GPS receiver technology and some care as to when and where
to operate, in order to minimize environmental error
factors. Recent advances in GPS receiver design have
improved resistance to multipath and RF interference, but
have not eliminated these factors completely.
Various Levels of GPS Accuracy
Using varying techniques, varying GPS receivers and
other equipment, a range of GPS accuracies can be
achieved:
_ An autonomous GPS receiver with Selective Availability
'on' will achieve a horizontal accuracy of 50-lOOm, 95% of
the time.
_ An autonomous GPS receiver with Selective Availability
'off' will achieve a horizontal accuracy of 12-1 5 m, 95% of
the time.
_ A code-phase GPS receiver using differential correction
techniques can achieve accuracies of between 0.5 and 3m, 95%
of the time.
_ A carrier-phase GPS receiver in kinematic mode can achieve
accuracies of between 1 and 5cm, 95% of the time.
_ A carrier-phase GPS receiver in static mode can achieve
reliable sub-cm accuracy. The important questions to ask,
when considering which GPS equipment is suitable for a
particular application, are:
_ What level of accuracy do you really need?
_ Do you need this accuracy in the field, in real-time, or
do you just need this accuracy when you return to the
office?
How Does Differential Correction Work?
Differential correction is conceptually a very
simple technique. A Base Station is placed on a fixed point
with known coordinates, and this Base Station then computes
differences between its known coordinates and
measured GPS positions.
These differences can then be matched up (on a
second-by-second basis) with GPS measurements from a roving
GPS receiver, and used to remove the systematic
(correctable) error factors. Differential correction can be
applied either in real-time or postprocessed mode.
The following diagram illustrates the differential
correction technique graphically:
Carrier-Phase Differential (Kinematic) GPS
The differential correction technique described
above applies to code-phase GPS receivers, which use the
transmitted GPS code information to compute pseudo-ranges
(distances) from the earth to the GPS satellites in space.
When a receiver operates in carrier-phase mode, it is
measuring a different GPS observable, namely the GPS carrier
wave. In order to obtain high accuracy with carrier-phase
measurements, it is necessary for a roving GPS receiver to
use information from a base receiver to compute the integer
number of GPS wavelengths between the roving GPS receiver's
antenna and the satellite(s). This technique yields
accuracies in the cm-range, and can yield mm-level
accuracies in static environments. In dynamic environments
(called 'Real-time Kinematic', or RTK), GPS is capable of
accuracies in the 1-5cm range.
What is Post-Processed
Differential GPS?
It is the reconciliation of the errors from
Differential GPS when the rovers return to the office:
_ No radio link between the base and rover(s) is necessary,
but
_ A base station must log data for post-processing with
rover data
What is Real-Time
Differential GPS?
It removes the need for post-processing by
transmitting differential corrections to the rover as they
occur, so that data is captured as accurately as possible
in the field:
_ Requires a base station computing differential GPS
corrections
_ Requires a communication link between the base and
each roving receiver.
Where do Real-Time Differential Corrections come
from?
There are an increasing number of real-time
differential correction sources, including:
_ Commercial Real-time DGPS providers, both Terrestrial
(e.g. RDS) and Satellite-based (e.g. Omnistar, Landstar)
_ Governmental providers, such as Coast Guard beacons
_ Custom systems, which require you to have a source of DGPS
correction in RTCM SC 104 format (i.e. a Base Station), and
a data link, for example, a data radio (modem and
transmitter) or GSM cellular telephone.
Almost all commercial sources of differential corrections
provide code-phase corrections only. The operational range
of carrier-phase differential corrections is currently quite
limited (50km at most, and more typically 20km), so most RTK
installations require you to establish your own base station
network.
Section 3 Differential GPS Application Areas
Differential GPS is now used in a wide range of
application areas:
_ Avionic navigation, including precision landings and the
aerial application of fertiliser & pesticides
_ Land and hydrographic surveying, including topological
surveys, seismic surveys and rig positioning
_ Agriculture, including crop mapping and the controlling of
harvesting machinery
_ Vehicle tracking, including in-car navigation, fleet
tracking and public safety applications
_ Precise positioning, including bridge sections, mining
machinery, construction and the general area of 'outdoor
robotics'
_ Mapping and GIS data capture, including in-field mapping,
GIS data collection and GIS data update
The following three sections focus on the GIS data capture,
vehicle tracking and construction application areas.
Section 4 Mapping and GIS Data Capture
GIS data can be captured from a number of sources,
including:
_ Digitizing from paper maps
_ Scanning paper maps
_ Traditional surveying techniques
_ Paper records & field notes
_ Photogrammetry
_ Remote sensing
_ GPS
GPS is a relatively recent addition to this list, and is
providing a cost-effective means of capturing GIS data in
the field.
How is GPS used for mapping and GIS data capture?
GPS-based GIS data capture systems combine three
technologies:
_ GPS for positioning
_ Ruggedised data loggers and field computers
_ Field-efficient data capture software
A GPS/GIS data capture system can sometimes be a completely
integrated unit, where the GPS receiver, antenna and
datalogger are combined in a simple hand-held unit. More
commonly, the three components are kept separate, providing
flexibility (and typically greater accuracy), where the
controlling datalogger can either be a hand-held, notebook
or pen computer.
What can a GPSIGIS system do?
GPS/GIS data capture systems perform three main
tasks:
Mapping and data capture
Mapping and GIS data capture is the process of
visiting geographical features of interest, and recording
both their position (using GPS) and any relevant attribute
information about these features.
Navigation
It is often necessary to find your way back to a
feature for the purposes of:
_ Regular inspection and maintenance
_ Repair
_ GIS update
GPS/GIS systems allow you to identify one or more
features in your GIS, load the position(s) of these features
into a datalogger, then navigate back to these features in
the field using GPS. In order to obtain accuracies of better
than 12-1 Sm in the field (or 50m, with S/A on), a real-time
differential GPS system will be required.
GIS Data Validation & Update
Once populated, a GIS needs regular update. Some
GPS/GIS systems allow you to take existing data into the
field. You can then use GPS to verify that you are at the
correct feature, before checking the status of the feature
and, if necessary, updating its attributes and/or
position.
For efficient GIS update, a GPS/GIS system requires:
_ A graphical display of existing data
_ Large data storage capacity
_ Real-time differential GPS
As such, a notebook or pen-based system is typically more
suitable than a hand-held computer for GIS update or
validation.
Advantages of GPS/GIS systems
GPS-based GIS data capture systems have a number of
advantages over other GIS data capture methods:
_ By being on-the-spot (in order to record a feature's GPS
position), it is possible for the operator to conduct
attribution and validation right there in the field
_ Ease of use. GPS systems require very little training in
order to reliably achieve good results, unlike conventional
(optical) survey equipment.
_ Accurate, efficient and absolute recording of
position. GPS positions do not rely on local landmarks, such
as kerb-lines or the position of boundaries, etc. Recording
systems which rely on measurements from such objects are
vulnerable to local change (such as a new road alignment,
the addition of a walkway, or the moving of all power lines
underground, etc)
_ Reliable positioning in remote areas. High GPS accuracy
can be achieved hundreds and even thousands of kilometers
away from any known or recognisable control locations.
_ Export direct to a GIS data format, with no manual
intervention, retyping, etc. GPS/GIS systems are entirely
digital, with data processing being highly automated. This
eliminates redundancy of data entry effort, and reduces the
likelihood of data capture errors.
Recent advances in GPS/GIS technology
GPS is a maturing technology, and a number of
recent technological advances have made the use of GPS for
Mapping and GIS data capture even easier and more
productive. These advances include:
_ Higher Accuracy. Typical instantaneous accuracies are now
around O.5-1.Om, and low-cost equipment can be used in
static carrier-phase mode to achieve accuracies
of<lOcm.
_ Better signal-processing allow use of GPS in more
"hostile" environments (e.g. forests, high-rise cities)
_ Wider availability of Real-Time Corrections enables more
users to operate with high accuracy in the field, which is
vital for navigation and GIS update operations. In addition,
recent GPS product releases have included integrated
GPS/DGPS receivers, which provide 'free' differential
corrections to users in some parts of the world, without the
need to purchase an external radio link.
_ Sophisticated Office processing software is allowing users
to maximise productivity in the office.
_ Integration with other technologies (e.g. laser
range-finders) allows GPS to be used in environments for
which it is not ideally suited by itself.
Section 5 Vehicle Tracking
In a wide variety of markets, GPS is being used for
vehicle tracking. These areas include:
_ Police and Military
_ Emergency Services
_ Fleet Management, which includes
_ Long-range Truck Companies
_ Ocean-going Vessels
_ Public Transport
_ Mining Vehicles
Fleet Management on a Construction Site
An area where GPS-based vehicle tracking provides a
wide variety of benefits is in the construction industry.
GPS can be used in despatch systems, to improve the
efficiency of vehicle use and also to provide audit
information (showing vehicle utilisation). In mining and
construction applications, vast sums can be saved by maxim
ising the utilisation of huge spoil trucks and concrete
trucks. In these applications, metre-level positional
accuracy is normally adequate, which ensures that the cost
of GPS equipment is modest. It is important that vehicle
tracking systems allow for flexible interfacing, as many
construction companies already operate communication systems
which can be used to carry GPS positioning information. In
addition to efficiency gains, vehicle tracking systems
provide enhanced site safety and security, where the site
manager can see instantly where all the site vehicles are,
and can re-route vehicles dynamically to meet changing
requirements.
Some large mines and construction sites, particularly those
in very remote and/or hostile locations, are investigating
the use of autonomous (driverless) trucks, which are
controlled from a central location and are navigated using
GPS. For this kind of application, higher GPS accuracy is
required (Scm or better), as is low position latency, so
that vehicles respond promptly to changes in direction or
position.
Trimble has recently entered into ajoint-venture with
Caterpillar to develop a full site management system for
heavy vehicles, which could be used in a range ofearth- and
ore-moving mining and construction applications.
Recent Advances in Vehicle Tracking Technology
PC-Based Control Software
Until recently, the central control system for a
vehicle tracking application required a workstation
environment in which to run. Workstations carry a high cost
to install and maintain, and workstation software
integration is costly. Trimble have now produced a PC-based
vehicle tracking suite called FleetVision, which
dramatically reduces the cost of establishing and
maintaining a vehicle-tracking system.
The FleetVision system uses standard software and data
formats, such as the Microsoft Access database engine for
data management, and the Maplnfo MIF format for the mapping
base. The use of standard software components allows for
easy customisation and integration with other software
systems.
Positional Accuracy
Many vehicle tracking applications do not require
differential accuracy; it's often adequate to know in which
suburb or even which city a vehicle is located. But some
applications do require higher accuracy, and this can be
delivered in the conventional manner (a telemetry link from
a base station to each rover), or (given that each rover is
already transmitting its position back to the control
center), an inverted differential correction method
can be used, where each rover transmits autonomous GPS
positions and additional GPS data back to the control point
which then differentially corrects this data and displays
the differentially corrected position. An inverted
differential system reduces the cost of the telemetry link,
which can then be one-way (from the rovers to the control
center), rather than having to be bi-directional.
Dead Reckoning
In some environments, GPS alone is inappropriate
for positioning. In vehicle tracking applications, best
results can sometimes be achieved using the combination of
GPS and Dead Reckoning (DR) equipment. GPS is used to
compute a vehicle's position whenever sufficient satellites
are visible, and the DR sensor (usually based on a
gyroscopic compass and odometer) is used to 'fill in the
gaps' when the vehicle is between high-rise buildings, in a
tunnel, beneath dense forest canopy, etc.
Integrated Communications
The cost of adding telecommunications equipment to
a GPS system can significantly impact the marginal benefit
obtained by implementing a vehicle tracking system. Recent
systems integrate GPS with telecommunications hardware in a
single package. Communications systems which have been
integrated with GPS include:
_ MAP27
_ GSM (using the Short Message Service, or SMS)
_ Cellular Messenger
Vehicle Event Monitoring
Some vehicle tracking systems allow events at the
vehicle to be transmitted back to the control
center. These events might include:
_ Door opening and closing
_ Vehicle loaded and unloaded
_ 'Panic' button for emergencies
The monitoring of events at a central control center can
allow the system manager to dynamically monitor the state of
each vehicle, and also of the 'cargo' (passengers, patients,
etc) being carried by the fleet of vehicles. Emergency
notification is also a significant safety factor in some
application areas (e.g. taxis and buses, which may be prone
to hijacking).
Section 6 Construction and
Engineering
GPS is increasing being used in construction and
engineering applications, in a variety of new ways. Specific
construction market segments which will be covered in this
section are:
_ Blast Drill Navigation
_ Earthmoving
_ Pile Driving
_ Structure Placement
Blasthole Drill Navigation
The navigation of a blasthole drill, and the final
positioning of the drill head using GPS, provides a number
of tangible benefits:
_ Reduced surveying needs. It is no longer necessary for a
survey team to stake out each drill site, based on a seismic
grid model. The model can be used directly by the drill
rig.
_ Reduced blasting costs with improved fragmentation,
derived from fitting the blast grid more accurately.
_ More even benches, providing a more uniform data model on
which to perform seismic analysis.
_ Production monitoring, where each drill site is mapped,
and the result (rather than the design or the stake-out
model) is used for seismic analysis, yielding more accurate
results.
_ Drill control. Robotic control of the drill head allows it
to be positioned more accurately than is typically possible
with human control.
_ Maintenance monitoring. Utilisation of the drill rig and
head is mapped and monitored, and maintenance schedules can
be automatically followed based on this data.
A pre-requisite for drill-rig installations is that the GPS
equipment and the controlling computer be fully ruggedised,
and resistant to shock, vibration, dust and wide variations
in temperature.
The system's drill map display shows the grid, and the
position of the vehicle, together with the drill head. The
operator positions the vehicle approximately over each
design point, and then allows the control system to
hydraulically position the drill head precisely on the
design point, before drilling commences.
Earthmoving
GPS is used in a wide variety ofearth- and
ore-moving applications, where it is fitted to dozers,
graders, scrapers and excavators. In all of these
applications, the piece of machinery whose position is
important is actually the blade, and the bottom of the blade
at that. This means that a complex set of rotary encoders is
used to translate the position of the GPS antenna (mounted
safely on the vehicle's roof) through one or more
multi-jointed arms to the blade itself.
The system then provides cut/fill guidance, in relation
to the required model surface. This shows the benefit of
GPS, as a fully three-dimensional positioning system.
Benefits of a GPS-based system are:
_ GPS eliminates the need for stakeout; a digital model can
be used directly to control the earthmoving machinery. The
elimination of stakeout crews provides a considerable labour
saving.
_ GPS greatly reduces the need for rework, which is often
due to hurried stakeout, stakes being moved or damaged,
etc.
_ GPS therefore provides greater machine utilisation, as the
machinery need never sit idle while waiting for a stakeout
crew.
_ The removal of stakes and stringlines allows for much more
convenient vehicle travel, as there is no longer a need to
work around these delicate physical obstacles.
_ GPS does not suffer from some of the disadvantages of
sonic & laser-based systems, which typically have a
short range and are suited only to simple flat or level
grades. GPS provides fully three-dimensional positioning,
and can easily handle complex grades and surfaces.
A recent advance in software has led to the automated
importing of designs into the setout or
earthmoving machinery. This eliminates the tedious (and
error-prone) step of translating a design from
a CAD system to the GPS-based system. Automatic importation
of design data addresses a number of
areas:
_ Intelligent selection of design features for
setting-out
_ Addressing horizontal & vertical geometry
_ Cross-sections
_ Templates
_ Design surfaces
_ User defined offsets
A MOSS road design can now be set out using GPS without any
user intervention, rework, conversion, etc.
Pile Driving
As with blast drill navigation, there are many
benefits from adding GPS control to pile driving rigs,
including:
_ Accurate guidance to the pile location
_ No stakeout required (and no human trying to remove a
stake from beneath a huge pile-driving rig!)
_ "As-drilled" information recorded for reporting
_ All weather operation
Trimble RTK GPS is at work in several deep-foundation
applications, where the operational constraints (accuracy
+or- 25 mm horizontal ) are very rigid, and the supporting
softwre is very sophisticated, providing:
_ Cab operator display device
_ Interface to tilt sensor
Real-Time Applications of GPS 15
_ Laser input to measure ground height
_ Automated final positioning
Structure Placement
Real-time Kinematic GPS is being used together with
sophisticated software to model the 3D position of a
structure, and control the precise placement of this
structure. Such systems require high GPS accuracy with very
low positional latency (it is sometimes impossible to 'bring
back' a structure if it is being positioned in a locking
fitment).
Recent examples include the Seo-Kang Grand Bridge, which was
erected across a river near Seoul, in South Korea. The
entire bridge span was pre-constructed, and was floated down
the river on barges before being positioned and attached to
abutments using GPS.
The project was managed by Hyundai Engineering &
Construction Ltd, with the culmination being the precise
guidance of a 150m single span road bridge section, which
had to be transferred by barge 2km down river then
positioned by anchor tension on supports and ballasting. The
project used specialised structure placement software
(called 'Target:Structures'), which provided a number of
on-screen information windows, in addition to a visual
indication of the position of the bridge section, in
relation to its design position.
Summary: The GPS Construction Site
GPS is being used successfully in all phases of
many construction and engineering projects:
_ Control
_ Topo
_ Setting-Out
_ Machine Guidance
_ Site Vehicle Tracking
_ Asset Management
The GPS Construction Site: An Example
A good example of a construction site where GPS was
used intensively (though not exclusively) by a large number
of sub-contractors, is the Vasco de Gama Bridge (NovaPonte)
in Lisbon, Portugal, where real-time Kinematic GPS was used
for:
_ Marine and land stages
_ Hydrographic survey
_ Dredging
_ Control survey
_ Topo survey/earthworks
_ Bridge pile positioning
_ Approach road setting-out
This example illustrates a key advantage of GPS: it provides
a common system which can be used through all stages
of the construction project.
Recent Technological Advances in GPS for
Construction
Multipath Mitigation and Removal
As mentioned earlier, multipath is the term for
signals which reach a GPS antenna via two paths (e.g. due to
reflection). Multipath can cause a GPS receiver to compute
inaccurate GPS positions, or to perform slow and/or
unreliable RTK initialisations.
The effect of Multipath can be mitigated by use of a
choke-ring antenna, but these antennas are typically too
heavy and expensive for convenient field use.
Recent GPS receiver software advances allow the removal of
multipath signals via patented techniques (Trimble's
'Everest' technology).
The mitigation or removal of multipath allows GPS to be used
in many environments which were once considered too
'hostile'; in many cases, these environments (e.g. mines,
downtown city streets, etc) are precisely the places where
the use of GPS can provide the greatest productivity
benefits.
Latency
In many machine control applications, it is at
least as important that a GPS receiver provide up- to-date
positions as it is that these positions be accurate.
Latency, or the tendency of positions to be made available
some time after the measurements on which they were based
were taken, is the cause is 'jerky' behaviour and
'overshooting' which is seen in many previous attempts at
robotisation. Low latency (l/10th or 1/20th of a second)
allows smooth automated movement of machinery and
structures.
| Abstract Of The Paper & The Profile of The Speaker | Speaker Index | Paper Title Index |
|
|
|
| |
