A different GPS Question

[Kolibear]

I have a GPS, so I don't want to know where to get a one, which one is the best or cheapest, nor debate the ethics of using it,

What I want to know is, how does the GPS second guess what you are going to do?

Assume that you are GPSing from A to B and on to C.

You get close to B, the Geeps is counting down the miles when you decide to cut the corner and route direct to C. So the miles-to-go to B slows down, stops and then starts to increase again.

How does the Geeps know that you are now en-route to C? At what point does it say 'Ahh sod it - this guys going to C'. Why doesnt it say 'Excuse me, you have to pass B (& collect £200) before I'll let you go to C'?

What would happen if you got a few miles out from B, Geeps is counting you down to C and you turned round?

[Scoob]

Tell it you are going to C, if you don't know how to do that you should read the instructions. If you go back to B, same as above.

[englishal]

Suppose it depends on the software, if it is not a particularly cleaver GPS , when cutting the corner, I would press the "Direct" o C button.....

[Aim Far]

The 196 seems to flip over to C automatically when you're on or near the track for the next part of the route. No idea what the criteria for that are.

[GroundBound]

I've written some software for an Air traffic Control program, which tracks a flight against its flight plan, using the radar track/plot (much the same thing as you are asking), in order to update the progress of the ATC flight plan.

The logic I applied was to identify whether the distance to the next navigatoin point was decreasing or increasing.

If decreasing, was the flight within a reasonable proximity circle (e.g. 1n.m. ) to indcate it had passed close to the point. If so, once the distance began increasing it was assumed the point had been crossed.

If the flight never reached a close proximity circle, then a check was made of the track (over the ground) and the relative bearing of the navigation point from the aircraft position. Once the relative bearing exceeded a parameter (e.g. 60 degrees) and the distance was increasing, this was taken as passing the point.

I would imaging the GPS does something similar, but with greater precision, perhaps.

(edited fro the last part of the question).
If you have passed (or been deemed to have passed) the navigation point, it is no longer the next point, so if you turn around, then you are flying AWAY from the next point - so go to point 2 above.

GB

[Aussie Andy]

There's a relatively simple geometric algroithm, from memory described in the user's manuals which can for example be downloaded from www.garmin.com I think... from memory it's when you pass through a line intersecting the two tracks in and out of the waypoint you've missed, if you see what I mean?

Andy

[dublinpilot]

Scoob,

I think Kolibear quite clearly knows how to tell his gps to go to C.

He's obviously already told it to go to B, and afterwards to go to C. He's simply interested to learn how the gps decides that he has already been to point B, even though he hasn't actually passed over that exact spot.

Telling it to go "directly to" isn't isn't always such a good option. Especially if you have a flight plan loaded into the gps to go to D, E & F after point C. Using the direct to C in this case, will result in that flight plan being disregarded, and therefore once he gets, to C, he will have to fiddle with it again to direct it to D, and again at E and so on. It's much better that he has done his flight plan on the ground, and don't have to bother with button pushing in the air.

Kolibear,

It's a question I've wondered about too. All I can say, is that it hasn't got it wrong so far, and always seems to know when I've decided to move onto the next point.

dp

[Circuit Basher]

Well, never having used GPS as a primary navigation tool, I wouldn't know!! ;)

As someone who has managed a number of somewhat similar software projects, what has been said by Ground Bound and Aussie Andy feels right - once you've approached the destination within a reasonable degree of accuracy and then your slant range to that point starts increasing, then the box will deem the waypoint to have been reached. Most commonly in the past, I was doing this during flight software proving for military avionics software releases. If you didn't get close enough, there used to be a multifunction button that permitted you to skip the waypoint and head for the next.

[2Donkeys]

I think Aussie Andy has already answered the question. The algorithm isn't a particularly difficult one. It is related to the algorithm that determines what lead an ILS GPS applies to warning you of the need for a change of HSI heading over a waypoint - the greater the angle of turn, the more the GPS anticipates the turn instruction.

2D

[S-Works]

In the Garmin units it is based on the automatic waypoint sequencing parameters (which are user definable) and angle of interception.

The unit will automatically advance to the next waypoint when you either pass over the previous, or pass with xMiles of it and then are heading direct towards the new one or are on an intercept. In the GNS430 manual and the 296 manual is a section on how this occurs.

[IO540]

Clearly there are several ways of determining whether a waypoint has been passed.

Sometimes the software does get this wrong. For example a KLN94B (an "IFR approved" GPS ;) ) has a bug in that if you enter a route which ends where it starts e.g.

EGAA ABC DEF XYZ EGAA

then, when still on the ground at EGAA, slightly more than half the time it thinks the first waypoint is XYZ when it "obviously" should be ABC. And no I didn't select reverse route. Honeywell confirm this is a known feature and offer a work-around like

EGAA ABC DEF XYZ HHH

where HHH is a navaid close to EGAA

:O

There is more fun to be had when bimbling around for a bit (away from the GPS route) and then [re]intercepting the GPS route. Most units don't do anything sensible.

There are other quirks, like the computed ETA to destination. This is based on the current GS, but that obviously won't be maintained if the route zigzags around a bit (the wind component will change). Same for any computed fuel reserve at destination!! Even with an air data computer available, the GPS doesn't calculate the wind properly for each leg.

One has to know the particular unit and its quirks. It's not a big deal because one has the moving map so one can always see where one is.

A lot of the programmers should be shot.

[Kolibear]

I asked the good folks at Garmin (as I've a Pilot3,) how close you had to get to a waypoint, and the response was that they weren't sure of the circle of proximity but thought it was about 0.1NM.

I think that they are out by a factor of 10 at least as I normally aim to get within 1nm of my waypoint and its always clicked over to the next one.

[tmmorris]

My AirMap 1000 was preset to 0.1nm radius for 'arrival' at a waypoint, but that might not be the same thing as passing a waypoint and sequencing to the next. I actually set it to 2nm: when it says 'arriving destination', for example, I want to look over the nose and see it, not find it's underneath me!

And anyway, who can fly accurately enough to arrive within 0.1nm (unless you never look out of the window, of course). (Indeed as a new GPS owner I can see the temptation of using it too much: I flew with it for the first time on Wednesday and became fixated for a while with maintaining exact track, but was woken up by a close call with opposing traffic (fortunately a nice controller had called it to me) and realised I should be looking where I was going.)

Tim

[Mike Cross]

Kolibear

On the GPS's I've had the changover has taken place when you cross the line that bisects the angle formed by your desired track towards the waypoint and your desired track to the next waypoint.

Easier drawn than explained.

Mike