iGuidance smooth driving ideas

PURDooM · 06-21-2006, 06:13 PM

Mods, I named the thread poorly, can anyone fix it?

-------------------------

Neoneddy and myself discussed an idea at the last minnesota meet. What if we could make iguidance display movement smoother? If we wrote software to feed iguidance gps data at a rate faster than 1 fix per second, would it update faster?

In theory, we could write software to read the NEMA data from the gps reciever, we could alter it before it got to iGuidance. For every update, instead of pushing the new point straight into iguidance, we could split it up into say 6 points that went from the old fix to the new fix and were fed to iguidance over the course of 1 second. But would iguidance refresh the map faster if it got faster updates from the gps stream? Can it even support faster com port rates?

I have to be somewhere in 10 min, so I don't have time to stick around, but I would be very interested in hearing your thoughts. I guess the first question is how do you test something like this... is xport open source? I don't know how to code a virtual com port.

xBrady · 06-21-2006, 07:33 PM

You can download the virtual port creator that Eltima has. www.eltima.com I think. Then just write a small app that sends NMEA data to the port that iGuidance is listening on. If you need a load of saved NMEA data I can send it to you. Then just pass the data faster than what would usually be sent, maybe all of it at once if you wanted. That's what I would suggest for testing. If iGuidance does support < 1 sec updates then I can maybe give some suggestions on writing an app to do what you want but I'm betting iGuidance won't let you go any faster than 1 sec.

Chris31 · 06-21-2006, 07:46 PM

I doubt the apps would accept higher update rate. Some apps do accept slower rate but not faster than 1 per secs.

By standard NMEA update rate is once every secs. The baudrate can be the standard 4800 or higher as long as the position update rate is fixed to one per sec.

I may have missed something but this is what I found out after some experiments.

jizzajam · 06-21-2006, 07:48 PM

The NMEA standard dictates that the gps device sends updates at 1 second intervals. also states that port speed is 4800baud. i doubt you could pump more than 2 nmea sentences per second through at 4800baud, and if you could there is no telling if the software would support it.

Wiredwrx · 06-21-2006, 07:52 PM

Ya, I think modifying the NMEA rate is not the way to do it. You have to mod the SW.

Michael

Marvin Hlavac · 06-21-2006, 08:19 PM

A quick and inexpensive test would be to just buy one of the 5Hz receivers, but my guess would be that this alone would not help. TomTom Go also uses a regular 1Hz receiver, and the beautiful fluid movement of the map is achieved by the software. We will have to wait till the Netropa programers feel about this one the way we do and implement this excellent feature into Intellinav software. Then we would see it also in iGuidance ;-)

briefnotion · 06-21-2006, 09:21 PM

I've been working on a similar app with Google Earth. Google Earth isn't really geared to point path movement, and when I was programming, I thought of, and gave up on the idea that I could move the screen slowly over one second. I was always thinking about it. Maybe I'll pick up on the idea when new tools are available or if I get more time and incentive. I'm happy with the way things are. Just have to think ahead before I make a turn.

This is what I figured out. You want the one second movement to position the focus on where you are now, and not where you where one second ago. Two ways, NMEA data is supposed to figure out where you will be when the next message is received by the turning rate and velocity. So drawing a point path to that location should be seemingly simple enough. Or you could work on the calculations yourself. Either way, it would be really cool if the screen scrolled smoothly. Good luck. And speaking of forum title mess-ups, I added a poll and there's no remove options.

Forum Link: http://www.mp3car.com/vbulletin/gps/77830-googleearth-gpsr-tracking.html
CommTest2 Link: http://briefnotion.250free.com/

xBrady · 06-21-2006, 10:32 PM

Quote: Originally Posted by briefnotion

And speaking of forum title mess-ups, I added a poll and there's no remove options.

Forum Link: http://www.mp3car.com/vbulletin/showthread.php?t=77830
CommTest2 Link: http://briefnotion.250free.com/

You should be able to close the poll if you want (I think you can close them if you made them). If it doesn't let you, let me know and I can do it for you.

briefnotion · 06-22-2006, 07:22 AM

It was fantacy. No lookahead that I know of exist in NMEA data. Sorry, must have gotten confused with waypoints. I supose us dreamers need room to be wrong, maybe put out a few possibilities. Yet still a regression alg should work. Now that I think about it, I may just mock up a version and see how it looks. I know it wasn't the intent, but thanks.

PURDooM · 06-22-2006, 09:14 AM

I don't need a regression algorithm, all I need is to delay the signal for 1 second and transition each position/heading linerally.

b8bboi · 06-22-2006, 09:36 AM

Quote:

all I need is to delay the signal for 1 second

I don't think this is a good idea. I believe the GPS signal already lags behind 1 second as is.

YellowBkpk · 06-22-2006, 11:21 AM

Quote: Originally Posted by b8bboi

I don't think this is a good idea. I believe the GPS signal already lags behind 1 second as is.

The NMEA position line that gets written out to the serial port is < 1 second old because it updates every second.

purdoom, I would imagine your idea would work just fine. It'd go something like this:

Remember the last GPS point
When the most current GPS point comes in, "draw a line" between the two points
Pick 6 points on that line, feed them to the fake COM port every 1/6th of a second
Set the most current GPS point to the last GPS point and start over

Wiredwrx · 06-22-2006, 11:25 AM

Quote: Originally Posted by YellowBkpk

The NMEA position line that gets written out to the serial port is < 1 second old because it updates every second.

purdoom, I would imagine your idea would work just fine. It'd go something like this:

Remember the last GPS point
When the most current GPS point comes in, "draw a line" between the two points
Pick 6 points on that line, feed them to the fake COM port every 1/6th of a second
Set the most current GPS point to the last GPS point and start over

The problem isn't ther NMEA sentence from the GPS, the problem is that IG probably only updates every second, and not when you send a sentance to it, but of course you/they can try. Ther other issue is can you send a sentence in 1/6 of a second.

Michal

0l33l · 06-22-2006, 02:48 PM

Quote: Originally Posted by Wiredwrx

The problem isn't ther NMEA sentence from the GPS, the problem is that IG probably only updates every second, and not when you send a sentance to it, but of course you/they can try. Ther other issue is can you send a sentence in w/f of a second.

Michal

So then how come yhou can pick a baud rate in iG?

Wiredwrx · 06-22-2006, 03:03 PM

Quote: Originally Posted by 0l33l

So then how come yhou can pick a baud rate in iG?

I thought the baud rate is just the speed the sentence is sent, not how often it is sent.

Perhaps you can get the whole sentence out in 1/6 of a second, but I don't think the SW will update that quickly.

I may be wrong.

Michael