Flutterby™! : GPS questions 2002-09-18 15:59:39+00

GPS questions

2002-09-18 15:59:39+00 by Dan Lyke 6 comments

Ages ago, Jeff and I talked about adding GPS data and a searchable map to the photo database. I thought having some data would be a handy place to start, so yesterday I made a list of intersections on the bus route to work, today I grabbed a Garmin GPS 12(that I think is actually Todd's), hopped on the bus with my Palm Vx out, ready to make annotations to that list... and... got no reception. Faraday cages suck. But that leads me to a few questions:

Is it normal for GPS units to have no reception inside commercial buildings or buses?
Is it normal for GPS to show wild swings of a hundred plus feet in altitude? Woohoo, now we're at -75 feet, now we're at 130 feet...

It'd be kinda cool to start to put freely available map data out there, and have data sets that we could play with, but only if it's not a hell of a lot of work.

Is it normal for GPS units to have no reception inside commercial buildings or buses?

GPS operates off microwave signals and require line-of-site with the satellites. This means no buildings, no vehicles, even along side cliffs (as you generally need a lock on four satellites). To get better, you need differential GPS - a GPS that uses the standard signal combined with a stationary beacon using RF or other EM bands to compensate.

My GPS (Garmin eMap) generally gets within 15-30ft using my external powered antenna. Consumer grade GPS only works to this level of accuracy. The differential GPS we have at work has sub-centimeter accuracy and is usable for survey work and such.

This brings up some stuff I hope to talk to you (and maybe Todd about) while I'm out your way next week. And to break the thread in a new direction, there is a big upheaval currently occuring in the Cartography world. As you may know, Cartography is the multi-millenia old science of making maps. In the past five years, computing power has reached a point to make Geographical Information Systems (GIS) cheap enough for wide-spread use. A GIS is essentially a SQL client tied to some graphics packages to allow you to link map data (or any visual representation of a land area) with other data. For instance, at work I can pull up a map of Chattanooga and ask it, based on recent census data, where the highest concentration of 18-year-old girls live. If I had another database, say a student directory from a local school, I could create a map showing where the 18-year-old girls are who go to that school - through a standard SQL JOIN statement.

This can be a wonderful thing but the problem is that the GIS industry is growing extremely fast (actually is in the top 10 areas of computer applications right now) and computer people are flocking to it without an understanding of the science involved in creating maps a presenting data in that medium. This is somewhat similar to the advent of desktop publishing (then crap newletters, now crap maps).

The biggest danger, however, is that this technology can be used quite effectively to assist in enviromental destruction (as well as protection), social abuse, marketting abuse, etc. The huge influx of poorly trained GIS people makes the traditional cartographers shake in their boots...

I still haven't been able to make heads or tails of my GPS unit, so I'm paying close attention here. But, um, Dan? How did "Nudity" get to be a topic for this thread? Was that just to see if we were paying attention? Or was it the up-and-down GPS version of your ride to work? :-)

Much of the fun and joy of Geocaching is dealing with wild swings in GPS accuracy when you lose line-of-site to the satellites.

Diane, I've no idea where "Nudity" came from. Damn, I hope some day that statement is taken out of context.

Eric, computer yahoos rushing into problems without domain knowledge, mapping is hardly distinct. And I like your comparison to desktop publishing; hopefully cartography won't deteriorate to the level design has. I would like to have a couple of mapping discussions because I think there's room for collaborative work ala CDDB (although hopefully something with a slightly more ethical approach will win) in mapping, and I'm interested in ways that I can search my data geographically and temporally.

dws, I have, at various times, nailed position to close to GPS accuracy with a good topo map and a $20 Silva ("Okay, take 5 steps forward, turn right, the trail should be..."), except that the topo and the compass become more accurate down in the valleys. Given the issues I was having walking down 7^th street this morning, I think it's a good thing that conventional navigation still be taught to anyone with mission critical needs.

Which sucks, 'cause I was hoping for a simple way to free the map licenses. Damn. Oh well, at least I need to get the serial cable for this one and explore the Perl GPS::Garmin module.

I can general images of maps for you from your GPS coordinates. I haven't looked, but there may be some opensource mapping packages that'll make the maps a little more dynamic. Let me know.

There really isn't much to the data that the Garmin acquires. It's really just XYZ data points with a time signature and maybe a label. Whether you export it from MapSource or dump it out the serial port, it's the same few bits. You then need to import that data into something that can map it. And watch out, the first time you map your data, you'll likely find that it's in the wrong hemisphere. 85 degrees west is about where Chattanooga is, which should be represented as -85 degrees in most of your mapping packages.

The danger with the lack of ethics in cartography doesn't really have to do with design - it has to do with exploitation. Pretty much all of the advances in GIS and Remote Sensing (aerial and satellite image interpretation) have come out of military and mineral exploration work. Maps can be very misleading and remote sensing data can be easily misinterpreted. One errant tech could make the Exxon Valdez look like a stain on a shirt.

What you mention about the topo and compass is the perfect example. What percentage of people would trust the results from a computer (GPS) over a manually read topo and compass? How many corporate execs would trust the results from a geek armed with lots of flashy GIS generated maps that don't accurately represent reality?

Dan. I don't think I got you a solstice present last year. Happy Solstice! enjoy it in good health. I think I remeber wild elevation swings and the trick was to leave the gps on and it starts getting more consistent over time.

Eric, I'll be dissapointed if we don't at least do dinner or lunch when your out here. Are you going to be out here for the Autodesk GIS exposition?