A 3.635Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE
The vision of real-time GPS as a public amenity or utility is not new, and has become a reality in many parts of Europe, Asia, and most recently the Americas. There are more than one hundred of these existing real-time networks. They are fully operational "utilities" providing an array of real-time and advanced GPS data services to their respective geographic regions. Some of these networks span entire countries, others represent regional cooperatives, cities, utilities, or are provided commercially.
There has been a lot of development and research done in the arena of realtime GPS networks, leveraging multiple reference stations and adding data-processing applications to generate network solutions for real-time observations. This was the next logical step in the evolution of real-time GPS; stepping beyond traditional single-base RealTime-Kinematic (RTK) surveying. Those involved in the network project being developed by Seattle Public Utilities (SPU) in Washington State reviewed the options currently available. S PU’s decision was to implement the unique real-time network solutions provided by the Trimble GPSNet suite of application software. The recent advent of networks of this type in the U.S. prompted Trimble to arrange for the first international conference of network administrators that utilize their GPSNet suite of products. In July of 2004, 69 attendees from 18 countries met at Trimble’s Munich, Germany location. The participants represented 33 of the more than 70 production networks of this type around the world.
The above German adage, "All things must be in order," sums up a lot of how surveying (and most things) seems to be undertaken in Germany. Well-known examples of this level of order exist in many German industries. The Reinheitsgebot, or Beer Purity Law as it is better known, has guided German brewers since 1516. While some of this `order’ borders on the extreme–like needing to take a test to get a fishing license, or to use a golf course–this mindset has produced some fine surveying instruments and software.
Particularly for the U.S. attendees, this first-of-its-kind conference was an eye-opening look into what has essentially been a paradigm shift in the world of surveying and mapping. Three days of training, presentations and networking with administrators from successful networks (each with several years in the business) left those of us in the U.S. contingent a bit awestruck. As the stunned Americans met at the end of each day’s proceedings to compare notes, it became apparent that our traditional view of real-time GPS was becoming obsolete, specifically in contrasting the terms adjustment, solution, and corrections. It was time to go back to school.
How does software help us do this kind of magic? To understand the Virtual Reference Station (VRS) solution offered by Trimble’s GPSNet suite, one has to step out of the traditional surveying view of error. Rather than merely adjusting the error out of a set of observations, which essentially often only spreads the error around, a network solution can be used to overcome the error. Advanced modeling, real-time access to source data, and the advantages of fully monitored reference can all but eliminate the PPM error we are so accustomed to dealing with in traditional RTK.
Typical sources of error in traditional RTK include instrument setup, atmospheric, and orbital elements. A base station network, stable mounted and continuously monitored, can eliminate the potential error we have every time we set up a temporary reference station. A live Internet connection can retrieve advanced projected orbit data, and this same network can generate modeled atmospheric data corrections. The GPS Net package generates a VRS set of corrections for the observer’s location, which provides all of the benefits of a reference station modeled precisely for your current field location.
This software is quite sophisticated and offers a wide range of real-time services (cellular data or radio) as well as other Web-based services. Along with the VRS type real-time corrections offered, the same modeling can be applied to single frequency work, improving Differential GPS (DGPS), or a host of post-processing options to include a type of "virtual" Rinex file generation, for instances when cell is not available. The G PS Net software also provides for continuous network monitoring. This includes coordinate integrity, individual station quality elements like multi-path, availability, signal quality, satellite tracking, and atmospheric conditions. Web applications display these statistics and reports to the user in real-time. Also demonstrated was a new "rover monitoring" application that can continuously test network quality from the perspective of a field rover or rovers.
There were presentations from many of the networks represented, showcasing a wide range of business models and some amazing user statistics. In short, the repeatability and reliability of the networks that were presented helped to erase the skepticism of the U.S. contingent, and quell the fears of the newest network administrators.
Among the most ambitious and successful is the nationwide network in Germany. Known as the Satellite Positioning Service of Germany (SAPOS), the nationwide network is comprised of regional and statewide networks, including data streams from some stations hosted by surrounding countries to complete nationwide coverage. Several of the SAPOS component networks presented topics such as network growth patterns, coordinate and rover monitoring, growth of user subscriptions, and studies by academic institutions on a number of system reliability elements. One particular study conducted by the University of Applied Sciences, Fachhochschule Mainz, presented by Michael Schöffel, compared system access initialization utilizing the Global System for Mobile communications (GSM) and General Packet Radio Service (GPRS) cellular data technologies. For the most part, Europeans are using the older GSM technology for their cellular access, and a majority expressed a strong desire to move on to GPRS. The irony of the U.S. being a relative newcomer to this "real-time revolution" is that GPRS is much more widespread here. That will make it easier in the U.S. to reap the benefits of these advances, such as the ability to implement standard user authentication like Networked Transport of RTCM via Internet Protocol (Ntrip), a public domain application designed to enhance all-important network security.
Participants presented an impressive array of reference station infrastructure solutions. Designs addressed ways to overcome extreme conditions in climate (Sweden and Finland), terrain (Austria and Switzerland), communications, power and bureaucratic hurdles. Hardware redundancy and remote operation of reference station elements were constant themes. The Austrian Positioning Service (APOS) uses Internet accessible power strips for remote cycling of receivers, whereas the Swedish Network of Permanent GPS Networked Stations (SWEPOS) houses duplicate receivers, power, and communications elements in a weatherized enclosure at each reference location.
The network hosts’ ongoing studies are addressing user confidence in the reliability and accuracy of such systems. Swisstopo, a Swiss nationwide network, has developed its own rover monitoring program the goal of which, as put by Simon Grünig of the Swiss Federal Office of Topography, is to "provide the best simulation of a typical rover at different locations." Typical results of a network test involve seven days of observations at a given loc
ation, connecting every 15 minutes for a one-minute location fix utilizing GPRS. Results analyzed included the all-important initialization times, and precision. A summary of representative results at 1* were 5mm horizontal and 10mm vertical with 97 percent of initializations OK. Impressive as these results are under such conditions, one can only improve such results if a local calibration were applied, as good surveying practices prescribe.
As expected, the typical markets for such networks include the surveying and mapping industries, which will benefit from the time/cost/accuracy improvements in both single and dual frequency observations. However, the list of further geomatics and navigation uses is expanding where these networks exist: area precise farming, land valuation, climatic research, maritime surveying, depths profiling, scientific and geodynamic investigations, construction surveys, deformation monitoring, and flood monitoring. Other uses include emergency and vehicle guidance systems, rescue services, traffic telematics, fleet management, vehicle navigation, utility registration, G I S, real estate cadastre, and land consolidation. Participants gave examples of many of these uses.
Two commercial networks were presented. GPSnet.dk, the commercial network spanning Denmark, is successful in offering the benefits of "network solutions," which is in contrast to a community base network in the same country that offers the traditional single-base form of RTK. In addition, one of the three competing commercial network service providers from Japan presented. The model in Japan is unique: a national agency maintains a nationwide network of 1,200 reference stations for geophysical research. A data stream from most of these stations is leased to the Japan Association of Surveyors, which in turn sells this data to three commercial operations providing the real-time corrections and a host of other GPS data services to the public.
Arguably, the most impressive benefit of this technology is the ability to enable grander schemes. Focusing on the cost benefit of such a network made it possible to fund FLEPOS network, run by the GIS-Flanders Support Center (a department of the Flemish Land Agency). This network covers half of Belgium, and is the backbone of a $100,000,000(US) comprehensive mapping campaign. The government and utility companies jointly fund this mapping effort, known as GRB. It is to be noted that the entire network infrastructure only represents 1-1.5 percent of this cost. Wim van Huele, project manager for FLEPOS withstood some good-natured ribbing from his peers at the conference who wondered what kind of hypnosis he employed on so many levels of bureaucracy to get the project funded. The U.S. contingent was particularly interested in his advice.
The proceedings (much to our relief) were conducted in English, though Spencer Reeder of the Spatial Reference Center of Washington (SRCW) gave a laudable introduction to his presentation in German. Spencer’s presentation introduced Washington State as a recent addition to the U.S. states participating in the National Geodetic Survey (NGS) program of Height Modernization. The NG S-administered grants enable states to improve the vertical control infrastructure and models utilizing digital leveling, gravity measurements, and G P S reference stations. An additional mutual benefit from these GPS reference stations could be data sharing agreements with regional and statewide real-time networks.
"The irony is that much of the world, especially the Europeans, have taken a U.S. developed technology and furthered its usefulness well beyond what we have done here," stated Reeder. "They have done this not for just a specific application but rather for a wide variety of users. The Spatial Reference Center of Washington, in cooperation with the Puget Reference Station Network (PRSN) collective, hopes to reverse this trend and bring this level of G PS usability to the entire State of Washington. This conference made it clear that such a thing is well within the capabilities of current hardware and software technology." Again, models that have been successful in Europe and Asia have done so with partnerships between the scientific, academic, and geomatic communities.
A highlight of the conference was a visit to the Bavarian Land Association (BLVA) network headquarters (a short, brisk, and very scenic walk nearby) to observe operation of a live network. The network monitoring and subscription tracking and billing applications were explained by giving plenty of ideas to representatives from the newly operational Puget Reference Station Network (PRS N). "I did not imagine that it would be so easy to administer our network," stated Roger Byarlay, surveyor with the PRSN. "Of course we get to benefit from the experience of outfits like the BLVA and others."
For those participants receiving administrator training, classes were held in the nearby village of Hohenkirchen. The unassuming structure in rural Bavaria is the workplace of small teams of developers. I must admit that the atmosphere of such an aggregate level of high IQs was a bit intimidating, and one would be surprised at the absence of pop cans and junk food wrappers on the clean wooden desks. It seems this clean and efficient environment is reflected in this amazing software.
Shortly after returning to the U.S., the PRSN crew held the official launch of the Seattle-based regional cooperative network. Seattle Mayor Greg Nickels was on hand to take the ceremonial `first observation’ with the new network. "This groundbreaking network will save taxpayers in Seattle and the region millions of dollars by reducing costs for surveys, civil engineering and construction projects," said Nickels. "The network will also improve response time to natural disasters and provide data for scientific research."
We can be confident of such claims from what we observed at the Munich conference.
Gavin Schrock is a surveyor and GIS Analyst for Seattle Public Utilities, where he focuses on using digital data to improve the cost ratios for engineering projects. He has worked in surveying, mapping, and GIS for 23 years in the civil, utility, and mapping disciplines. He has published in these fields and has taught surveying, GIS, and data management at local, state, national, and international conferences.
A 3.635Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE