The entry of generalised software companies such as Google, Microsoft and Oracle into the geospatial information systems (GIS) market will certainly shake it up, says Intergraph CTO Peter Batty. There are already some efforts to go head-to-head with the giants, for example California-based GIS specialist ESRI’s competitor to Google Earth. “This to me seems a tough strategy to pursue,” he says.
Intergraph and probably a number of other GIS companies will seek to complement the mass-market offerings with their own “niche” additions. “Those companies don’t do all the things we specialists do; basically they publish and distribute information.”
The supply of the information and its further analysis and correlation with other databases “for example, mapping the patterns of occurrence of crime in a city, or maintenance of the electrical network” is an area where the specialist companies will still dominate, he says.
Batty, respected as one of the leading minds in the geospatial field, spoke at a one-day seminar organised by the Geospatial Information and Technology Association (GITA) in Wellington last week.
The change in the marketplace will accelerate a movement common to many areas in ICT, towards a concentration on specific applications rather than technology. Like e-commerce, GIS is simply becoming part of the way business is done rather than a dedicated discipline.
“We don’t have ‘numerical information systems’ conferences any more.” At the same time, the GIS experts can take advantage of the greater interest generated by the giants in the value of geospatial information, he says.
These companies are also “changing the way we do integration”, through such general-purpose tools as BizTalk, BPEL and Oracle Integration Tools, as well as the more general concepts of web services and service-oriented architecture. Integration of different technologies and databases is a necessity for fully productive use of GIS and it is becoming easier than it was.
One of the most marked movements in GIS itself is the maintenance of information in real time. The databases of, say, a utility company’s equipment can be updated in real time by the maintenance engineer simply pressing a few keys to record that he has installed the new transformer. Increasingly, through RFID markers and proximity detection, there may be no need to enter any data manually; a sensor on the pole will record automatically that new equipment has been installed.
Security and location detection is another increasingly important and lucrative use of GIS, Batty says. If analysis of the inputs from sensors and cameras around a building uncovers a suspicious pattern of movement, trying a number of doors in sequence for example, then guards can be sent quickly to the site of the apparent problem. Intergraph is one of the suppliers of a huge security and movement tracking system for the New York subway.
The rise of wireless communication is a boon to real-time recording of geographically distributed data and will hugely accelerate the trend to such use, Batty says.
Such huge amounts of data, some of it historical, raises problems of data quality, currency and accuracy, he acknowledges. Where the data comes directly from the real world, for example aerial photography, “it’s easy to make it reliable and consistent,” but for historical data in the form of maps and lists, “I don’t know that there’s a particularly easy answer.”
He suggests, however, a variant of the open source advocates’ “many eyeballs” theory; the more people that use the data the greater the chance that inaccuracies will be detected and weeded out.
Standards for geospatial information are in an uncertain state. “We have an Open Geospatial Consortium (OGC) that has created interoperability standards supported by the existing industry players, but along come Microsoft and Google with their own idea of standards, and we have potential conflict.” There are workarounds, he says; developers are putting “wrappers” around Google’s KML language to accommodate it to OGC, but the long-terms future is still difficult to predict.