Guest Editorial: GIS: The one that almost got away
Professional Surveyor Magazine - January 2011
by Richard Trainer
Members of the surveying profession have been talking about GIS for years, but the industry has missed the opportunity of GIS data collection for a wide variety of reasons. In this article I review a bit of history of the surveying-GIS relationship, make lots of observations, participate in huge generalizations, and try to discover where surveyors went wrong and if it’s still possible to improve this relationship.
Surveyors should have been involved in GIS right at its beginning, and there should have been a partnership between the two industries. The most famous attempt was the Leica
partnership that ended about 10 years ago. Leica used ESRI technology to produce a system that was perfect for the Swiss cadastral surveyors. Unfortunately, only a microscopic segment of the GIS market could ever be interested in this system, and even Swiss cadastral land surveyors aren’t particularly interested in Swiss cadastral surveying systems. So this opportunity was lost, along with many others. What are the surveyors who designed these and other systems not understand?
Surveyors don’t understand that GIS requires a wide range of accuracies. In North America when you talk about GIS, most surveyors think of low-accuracy applications, and in their minds this isn’t real surveying. Sure, quite a few application areas need only low accuracy, but many of them also require precision. For example, a water company might need its manholes positioned to only a foot, but they always need the cover level to an inch. In Europe things are different because the majority of GIS users need precision accuracy for everything. Yet, even in Europe, the GIS world has not welcomed surveying methodology with open arms.
Surveying data collection methodology is antiquated and not fit for purpose. Surveying suffers from having computerized too early. When the first total stations with data loggers appeared in the 1970s, the only thing you could enter with each point was a number because that’s all the computer technology could handle. Data loggers advanced to using alpha codes, and then these point codes became feature codes, and feature codes had attributes attached for GIS. Are all these codes necessary? If you are surveying a tree then why not just say it’s a tree and put the attributes on it? You cannot expect GIS “surveyors” to have to cope with surveying methodology that is archaic and should have been replaced years ago.
Surveying instruments are unnecessarily complicated. Currently on my desk is a total station. I won’t say which one, but the analogy I’m about to use works for many of them and applies just as well to GPS. This instrument has a keyboard on the front, a keyboard on the back, a complex multi-tiered menu system, and about 100 possible settings, any one of which, if set incorrectly, will cause it to malfunction. I had to call for help with this total station earlier today, and the attitude of the help line was that I should have understood the problem and it was my fault. Well, a total station is just measuring two angles and a distance; that’s all it is. If the manufacturers can’t make two angles and a distance simple, then it is their fault and not the user’s. A GIS user will not put up with this, and frankly neither should surveyors.
GIS is encroaching on the traditional surveyors’ markets. As GIS technology improves, the GIS system is getting closer and closer to the field. Originally GIS fieldwork was done just by logging GPS points directly into the GIS, but with the advent of metadata a whole surveying data structure can be replicated in GIS. Cadastre and land registration is really a GIS application; right now the GIS is office-based but it could very easily expand into direct field data collection and into competition with many surveying systems. Utilities are also big GIS users that are rapidly expanding their GIS into field precision data collection. If you take cadastre and utilities away from surveying, the only major market left for surveyors is construction, and even that isn’t safe from the encroachment of GIS.
The GIS industry has done more than adapt some surveying techniques; GIS is starting to take over the surveyors’ toolbox.
For example, let’s take the rather excellent Topcon
GRS1. Intended for GIS users, the GRS1 connected to a VRS network gives the GIS user a handheld unit that can deliver centimeter accuracy, and if the software is designed correctly, the user just needs to turn the power on to get that accuracy. The user doesn’t need to know about base stations, traversing, resection, or even to some extent coordinate systems. So why is a surveyor needed any more?
Surveyors can still get in on the GIS action. The surveying profession might be wounded by the upstart GIS industry, but with a slight change of thinking surveyors can still provide vital services to help GIS users. The one key area that GIS doesn’t understand is precision and the proof of that precision. The ESRI user conference actually holds a competition for who can produce the prettiest dataset. The last time I was at the conference there were hundreds of entries, but not one of them even mentioned how accurate the data was or in most cases how it was collected. Apparently in GIS, good data equals pretty data.
A surveyor knows about the accuracy of data and how to prove it. It’s interesting that the United States is the most litigious country in the world, yet most of the GIS data collected in the states is not supported by any information on how it was done and how accurate it was. This is what surveyors can do for GIS.
The survey industry had its chance to be in at the beginning of GIS and has failed. Now the GIS industry is absorbing the techniques and even hardware from surveying and is in a position where it can perform its own data collection with its own methodologies. It is also doing so in vital surveying markets.
The key for surveying’s survival requires a change in thinking on the part of surveyors. They have to change some of their archaic methods, understand that not every application needs millimeter precision, and most of all stress the need for verification and proof of results. It is still possible to fix the GIS-surveying relationship, but time is running out.
Richard Trainer is a designer and programmer of the Cartogoo handheld GIS data collection system, and has worked in both the GIS and surveying industries for many years. He can be contacted at firstname.lastname@example.org.
» Back to our January 2011 Issue