The Seriousness of Bad Street Data
Written by Chad on November 17th, 2008Ok, so everyone knows how I feel about bad mapping data.. mainly Tele Atlas (lesser extent: Google and it’s bad choices). But listing to my fire pager and scanner.. I am seeing a real concern with just how badly some map data really is.
Emergency Responders can’t get to some locations because the map data they have is WRONG. As I am writing this post, I have listened to two calls where the responders can not find a location, the 911 center is telling them where it is, even said they looked it up online and still, the responders are not at the right location. For the second call, they finaly gave up and got out a county map book and looked up where the address was and found out they were 3 miles from where they should have been.
That kind of a mistake really is the difference between life or death. I do not know what the local 911 center uses for mapping locations (I’ll find out when I take the higher level NIMS training) but going by the fact of the road locations he was giving was not matching with what it really was.. it could be tele atlas data.
Now luckily, 97% of the call outs one would get, you are going to be in your local area for the call. But we do have Mutual Aid (if needed, we go to other communities where their fire department needs assistance) and you could end up, as a first responder, in a location you are not familiar with. All the trucks in our fire comapny have paper maps for the entire county, and the counties to the east and west of us.
19
PM
While this is a problem, it is becoming less so over time.
I recently worked with one of the larger companies providing CAD (Computer-Aided Dispatch) software, providing training support for, among other things, the mapping component of the system. The process of getting accurate GIS data in place is not simple, but we (and most all vendors) recognize its importance and devote a lot of resources to it.
Starting with the pre-sales engagement, we work with the local authorities to generate “base map” data. The GIS package includes layers for everything from terrain to streets to fire hydrants to structures to electrical services to, in some cases, structure occupants and owners.
Our engineers then work with the client to verify as much of that data as possible, to develop the initial base map set for the CAD installation. Fully 70% of the pre-sales activity is involved with compiling and verifying the GIS data, usually to the highest GPS coordinate accuracy allowed at the municipal level.
Once the sale is made, pre-installation training is provided. About 60% of that (several hundred hours per person) is devoted to training muni GIS staff on maintaining, updating and enhancing the GIS data. Finally, once the system is deployed and active, our company provides 24/7 support on all issues, including support for near-real-time GIS updates. The entire system is designed to permit the pushing of GIS data out to field units as rapidly as possible. In many cases, first responder corrections can be incorporated into the main GIS database and be pushed out to second-responders before arrival. So things like temporary road closures due to traffic accidents are logged into the CAD system, which updates the GIS data in near-real-time to alert responders to other incidents that that route is unavailable.
Properly executed (and that is a major caveat), the mapping accuracy is constantly being updated. All city or state services can be involved, from the planing and permits departments to the motorcycle cop on the street. CAD data is constantly changing to keep up with reality, and the better syatems are specifically designed to be both self-healing (coordinate data is constantly checked for conformity: if a fire hydrant is reported at GPS coordinates that would place it in the middle of a house, for instance, it’s flagged) and comprehensively maintained.
So, things are getting better. Of course, this costs money, and not every state, county, city, town, parrish or whistle-stop can afford the most advanced systems. But pretty much every US state, county and city of any reasonhable size does have this technology, and the maps are getting better and the GIS datasets are getting deeper.
[Translate]
20
PM
I think the trick here is to make locally created and provided data the authorative GIS source while still X-ref’ing with open source data.
Here (SW Montana) the local Public Works department has created most of the underlying GIS data on streets because of its responsibility for both water and sewer infastructure and road condition and signage.
The emergency responders have access to that GIS data as well as the TeleAtlas and ESRI Streetmap stuff and are able to query for differences in the data.
Most discrepencies are now common knowledge and because there are only a few, it is bacically a non-issue here.
Note that I am talking only about the county level (and the smallest county geographically in the state, with only around 40k residents.) I’m sure the issue gets to be a bigger problem state-wide or even in counties without such already integrated GIS systems and experience.
[Translate]
25
AM
Emergency responders need to be familiar enough with their local geography that they don’t need GPS. It should be part of their training. They should know all the major streets and best routes between various areas of the city depending on the time of day. And they should have paper maps so that somebody other than the driver can be looking up more obscure streets while they’re enroute to the general area.
[Translate]
25
PM
Haha. That is a good laugh
That “might” work in an urban area, but I can tell ya.. in a country/rural area, you can live in an area all your life, and STILL not know all the roads around there.
And you can even come to roads that are not even on any maps (I have found local roads that are listed on topo maps, but do not show up on any other maps).
[Translate]