Route points - placement accuracy.

[Mikey]

An aspect of BC route building that has puzzled me since I began learning its use (still learning) is the accuracy of route point placements on the route map. By accuracy I mean route point placement location ref the magenta line.

I assume the Garmin road routes are positioned on a centreline that joins a myriad of coordinated points. I think I have read that, are these ‘Ghost points’ btw? I also assume the route point placement (centre of the dot) has to be within a specific but unknown distance (L and R) of the road centreline or, in the case of a Via point, the XT might demand a U-turn.

There must be an allowable error +/- of the centreline for route points and I wonder what that is. Is it always sufficient for the route point to be ‘within’ the magenta as my eye judgement perceives or is there a possibility of a hidden, unexpected, inexplicable, RUT demand in this? That is, can the route point placement look okay to me but XT thinks otherwise?

Thank you, Mikey

[Peobody]

I don't know if there is is an exact distance from the route point coordinate to the proximity where the XT considers the point visited but my experience puts it around 30 feet.

There considerations for route whose purpose is solely navigation and route points that are for planned stops. Navigational route points are placed on roads that you want to traverse. It is important that they be place precisely on the road, and in the correct lane if the road is divided. You don't need to do this at first placement, but once you have the route complete, double click on it to open its details and tick the "Center on map" box at bottom left. Now click on each one of these points to check their location. They will be map centered. Zoom in as tight as possible. Enabling the "Move" function then hovering over the route point helps show its location and allow you to slide any off-road point onto the road. Make this process part of you normal routine when finalizing a route.

As for planned stop points (Via points), this is tricky. Keeping the 30 feet proximity distance in mind, many POI coordinates are placed such that there is no way for you to get your XT close enough to the coordinates for it to be recognized as visited unless you carry the device into the building. This is a problem for many restaurants, hotels, museums, large fuel stations, and the like. There are a number of tricks for dealing with this. Mine is to place a route point on the road just prior to the expected entrance. This insures that the XT will detect that I have visited the route point. If point is place that I will be stopping at and then continuing on then I will move its POI point to the road. Overnight locations require a different thought process. On arrival, you really don't care if the XT considers the point "visited" since its the end of the day, but on departure, you can't use a hotel POI as the starting point because where you parked may be far enough away from the POI coordinates that the exit will route you towards the hotel office. In this case, place a route point on a road in the direction of travel outside of the hotel parking area. This applies to the starting point of any route. Place it away from the physical starting point. It is possible to create a loop route that looks good on the map but the XT will tell you that you have arrived before you have gotten started. All it takes is a loop route using non-alerting shaping points and start and end points as your physical starting location. As soon as you start moving, the XT will sense that you are at the end point and tell you that you have arrived. Been there, done that, got the t-shirt.

Regarding the RUT issue, a misplaced route point can force you to have to tell the XT to skip that point. Doing this can then lead to RUT if the route has not had a RUT treatment applied.

[rbentnail]

I don't know if there is is an exact distance from the route point coordinate to the proximity where the XT considers the point visited but my experience puts it around 30 feet.

I'd say that's about right. I've had "missed the point behavior" happen twice to me recently, once on a 6 lane hwy, apparently I was in a lane too far away; and once in a medium-to-large parking large lot.

There considerations for route whose purpose is solely navigation and route points that are for planned stops. Navigational route points are placed on roads that you want to traverse. It is important that they be place precisely on the road, and in the correct lane if the road is divided. You don't need to do this at first placement, but once you have the route complete, double click on it to open its details and tick the "Center on map" box at bottom left. Now click on each one of these points to check their location. They will be map centered. Zoom in as tight as possible. Enabling the "Move" function then hovering over the route point helps show its location and allow you to slide any off-road point onto the road. Make this process part of you normal routine when finalizing a route.

Good description, I do this with every route every time. Many times I also check the route again after recalculating following a map update. I very rarely have any planned stop via points but when I do, to avoid problems, I make separate routes with the planned stop as a destination rather than a route point. Simple naming convention helps me keep them in order as I'm not so ocd as to require multiple route to be listed in order .

[Mikey]

Thank you Peobody for your detailed reply and time taken. I understand all you have said. Thanks to rbentnail too.

Mikey.

[jfheath]

I agree completely with what @Peobody has said.

I'll add a few snippets.

• There is some indication that from the 595 onwards, that the Zumos have been applying a certain amount of 'Fuzzy Logic' in defining route points - perhaps to cater for the increasing number of people who are using other route planning software. The maps used in the variety of mapping programs, differ, so an accurately placed point on one map will be in a slightly different location on another map - ie not on the road. So I reckon that what happens is that the Zumo takes your route point coordinates and finds a place that is nearby that it knows and has a name for. It then relocates your point to the one that it has found and it using the 'nearby' coordinates that it has in its database. So essentially, your point has moved and it now has a new name.
  
  With the XT1, this behaviour can be prevented by creating a Waypoint and using that in the route - set as either a Via Point or as a Shaping point. It then keeps its name and its location.
  
  With the XT2 the same is 'almost' true, but the XT2 manipulates the route in other ways - eg if you set your route point as a Via, it may move the via point slightly, but it will not move the waypoint with it. The result may be that your route does not pass through your Waypoint at all.
• RUT behaviour - I have only ever observed this after an imported route (from a gpx file) has been recalculated after pressing skip. Other people say they have observed it at other times. But I have not been able to reproduce what they described. RUT behaviour is very specific and it involves the nature or the behaviour of the route being altered by the Zumo. Once this has happened , the navigation behaves very much like a 'Track-Trip' - which has no route points at all and not like a route which navigates from one point to the next. A Simple way to avoid RUT is to never press Skip. Or if you have to, then stop and re-load the route and select Closest Entry Point.
• When is a route point regarded as being visited ? I do not know. I do know that Via points which are placed even slightly off road are not regarded as being visited. 30ft into a field ? I suspect not. Shaping points - they don't care as long as you rejoin the magenta line after going past them. But in Basecamp, if when I zoom in on each point I see even a tiny spur leading off at right angles - I fix it.
  
  But this example is interesting: On one of my test routes, I placed a Via Point. A couple of hundered metres before that was a left hand junction. I decided to take that left hand turn and not visit the via point along the main road ahead - expecting it to start complaining and demanding that I turn bakck and visit it. It didn't - it just calculated the new route in the direction that I was heading. Odd, I thought.
  
  So I went back to see what was happening. After a few trail runs with the via point moved further and further away from the junction but still on the main road - I discovered that providing that the satnav had said something like 'In 300 meters arriving at Via Point on the left', then it considered it to have been visited. If I moved that via point further up the road so that I had turned off before it had made the announcement, then I had to go back. I think that is about 3/4 of a mile when it first announces.
  
  In practice, this is rarely likely to happen. If the Via is placed to the side of the road, in a field, then the Zumo will have to make an announcment to turn into the field before it can announce that you are about to arrive at the Via Point. Places where turn instructions are given will appear as white arrows on the map. The "arriving at Via" announcement can only be made after that instruction to turn has been obeyed.
  
  So I guess the limit is within the bounds of what it allows for a road. On motorway slip roads, it doesn't even notices that I am not on the motorway, until I turn away from it. I am near and heading in the correct direction, therefore it 'assumes'. How much leeway does it allow ?
  
  I don't know. But clearly less than the discrepancy of road positions between different maps.
  
  Have you ever plotted a track of your route and transferred it onto Google maps ? You might have to convert it to KML format which Basecamp can do. It's been a while since I did this - but the results on twisty roads were less than impressive.

[Peobody]

In thinking about my 30 ft estimate, I base that on a single experience at a fuel stop where the pumps were a good distance away from the associated convenience store. I fueled up and as I was about to pull onto the road I noticed the XT still showed directions to that stop. I turned around and rode close to the building which satisfied the XT. I don't know where the actual POI coordinates where in relation to the building. I also don't know the exact point where the XT recognized 'arrival'. The 30 ft is an estimate of the distance from my closest point to the center of the building. If the coordinates were the front door then the distance may have been much less.

[smfollen]

I too agree with everything above. I'll take it just a bit deeper for two reasons: (1) accuracy and (2) because the original post asked about ghost points (aka hidden points).

Accuracy:
Maps, and the roads or paths shown on them are of course digital (or paper) representations of the physical world. They are impressively accurate, but not perfect - that is a route point position, lat and lon, obtained from a map may not exactly indicate a point on the actual physical road. It will be close, but may include a small error.
Similarly, gps receivers, like the zumos, determine our position as we move along with a certain accuracy which can vary by a few meters / several feet.
Further, although we tend to think of our "GPS" (SatNav device) continuously knowing where we are, what it actually does is calculate point positions at regular frequent intervals - that is, our position is a series of dots, not a continuous line. The faster we are moving, the further apart those dots will be. How close we get to a route point depends on the net sum or difference of these 3 inaccuracies.
... and, as @jfheath points out, the zumos seem to attempt to applies corrections based on where it "thinks" the road actually is.

I think the best we can do, as @Peobody said, is to zoom in on the map and place the route points on the indicated road as accurately as possible. That generally has worked very well for me.

Ghost points:
Zooming in very close on a map reveals that the roads are made up of series of points connected by straight lines.
When we specify a few route points, via and/or shaping, and then calculate the route, the software finds an appropriate path over the roads between each pair of route points. The points on the map representing the roads that path follows become the ghost points. In a route in a gpx file, they are the route point extensions. If a track is generated, they are the track points.

So, yes, I believe the points representing roads on the map, the "ghost" or "hidden" points of a route, the route point extensions and the track points are all the same points, just named and formatted differently.

[Mikey]

Thank you everyone for your informative thoughts and insights into point plotting accuracy.

I am gradually gaining confidence that points plotted on BC maps ‘to the best on my human ability’ and located wisely will generally provide for a successful route and trip.

On a recent 3 day ride (3 legs - 2 BC Imported, 1 Saved direct to XT), the 3 trips functioned perfectly, even with the numerous Via points I had included. These trips will be ‘keepers’

In consideration of keeping a Trip what is the best way to store these successful trips ? Note; I changed one of the Imported trips to a Saved trip on my XT (one nearest my home) just as a trial of the Imported to Saved process.

Thanks again, Mikey.

[smfollen]

@Mikey Glad to hear it is working for you.

You can save your routes as gpx files, as you just did. You can keep some of those files on an sd card in the zumo for easy access. You can copy or move them to a personal computer or similar backup storage. You could use Basecamp or Trip Manager or Windows file explorer (or the apple mac version) to do that. You could also use Trip Manager to save the trip files directly to a PC but gpx files are more universal than trip files.

The zumos also log tracks as you travel unless you turn it off, so you can save your logs or "unsaved tracks". @jfheath has said he does that for all of his rides.

[lkraus]

I keep all my planned routes in Basecamp, along with a track created from the route. After the ride, I also store the recorded track in Basecamp, named with an "AR" suffix (As Ridden). This helps me see where I might need to change the route based on the actual experience.

Be sure to recalculate your stored routes in Basecamp before loading them to the Zumo again - maps may have changed since you last used the route.