njkayaker

Many people have no idea that a map can be incomplete or wrong.

I keep an eye out for discrepancies and, if I find any, I update the Openstreetmap data.

The newer Garmin Edges come with maps based on Openstreetmap data.

It's not likely that there will be significant differences between the files produced by different route planners.

There might be different features (not all of them can produce course points) or there might be differences due to using different maps.

Probably not (except through using different maps).

The Garmin Edges (and, I believe, all of the units used for cycling) use a synthetic track (something that looks like it was the result of recording a ride but with fewer points in straight sections).

RWGPS produces TCX files and GPX track files. These contain the same track points (both are track files). They won't work differently (they contain the same data). The TCX files can contain course points, which supports a separate feature. (Note that the Garmin Edges convert TCX and GPX files to FIT files.)

The "GPX route" produced by RWGPS just contains the few "must pass through" points. These show up on the Garmin Edges as connected straight lines. The Garmin Edges can't use these files. To use these files, the GPS unit has to calculate a route between the "must pass through" points.

Iride01 

That is curious, and though I see posts about cycling GPS's taking users the wrong way on streets in the UK, particularly around traffic circles, I suppose that to be wrongly applied corrections from the map data bases and programming for the differences compared to what side of the street we drive on in the US vs some other countries.

So what's up with this opposite turn? Is it only at a particular location? Is it random? Does it happen every time for that location? Bike trail intersection? Motor vehicle roadway intersection or crossing?

What happens if you follow that turn instruction? Assuming there is a road or path.

Seattle Forrest

As a hiker, I see this sometimes. Get off the highway, follow some local road a while, then a faint dirt road to the trailhead. The car GPS shows me on the paved road for a few hundred yards, then it accepts that I really have left the road, and shows my actual position.

Side note: I have a cheap electronic bathroom scale. It's also got some cheat programming in it. If I step off and back on, I get the same number. If I step off, pick up a half pound glass of water, and step back on, I get the same number.

I suspect in both cases it's that the sensors aren't perfect and the data is noisy and this is how people thought to deal with it and provide a consistent user experience.

njkayaker

It's hard to explain what might have caused that with a route I can't look at.

With RWGPS, the default course point (cue sheet item) type/icon is a left arrow. So, some wrong left turns might result from that.

Also, in RWGPS, the course points (cue sheet items) don't get recreated/updated when changing routes.

I'm not sure if that explains the problem Koyote is talking about.

But the Edge GPS units can't really determine heading if you are stopped (they don't have the compass necessary to do that).

I have seen issues related to wrong headings. Moving a little usually clears those up.

njkayaker

The Garmin Edges have a "lock on road" option to enable this behavior.

I'd rather see the actual location because it lets me know that the map might be a bit off.

I'm also not bothered by the difference because, if I'm riding on the road, it's easy to be aware that I'm on the road.

Ideally, the map draws roads aligned to the centerline of the real world road. A GPS mounted in a car is generally fairly close to the centerline. A GPS on a bicycle is often 10-14 offset from the centerline.

njkayaker

It can happen if the route isn't cleanly drawn. Sometimes, the track has little deviations that the GPS unit interprets as a turn.

If using RWGPS, it can be because the default cuesheet type (left) was used or because the automatically-generated cuesheet items weren't recreated when editing the route.

In rare cases, it sometime just happens (I think I've seen it on the 800).

It's generally not a problem (for me) because I look at the map (and can see/follow the track line).

Iride01 

[QUOTE=njkayaker;20306599]We'll just have to disagree then. No bigee, I'm up for a beer, coffee or what ever legal substance you relax with during conversation.

Part of what I meant to say about trying different sites was to include that for some, the user interface just works better for them. The "intuitive" tools at one site are not necessarily "intuitive" to all persons equally. Back when I was messing with courses, I never could get Garmin Connect to generate a decent course of the routes that I ride locally or wanted to explore. RWGPS tended to be better than Garmin or Strava for me at that time.

I also disagree that a .tcx file from RWGPS will be identical in data to Strava or another, some sites I have to do a lot of pinning and rubber banding to get the course to go where I want. Other sites much less for the same route. But, I've never opened the files up to compare them side by side. They are only text files so anyone can look at them with a text editor or word processor. They are in a mark up language format much like html. Maybe someone will get interested enough to investigate for us.

Also, even if the files are the same, the navigation software is different even among the various Garmin Edge devices. Some of the Garmin's, at least from what I've gleaned in the Garmin Forums, will interpret the course file rigidly. Others will try to apply that course file to the map the device contains. (the course files never contain a map) and then build your turn cues based on it's internal programming. Some users say some models even route them on other roads. All this with other users of the same device in other localities not claiming any issue what so ever.

As for OpenStreetData, it's a great resource. More accurate for some, but in my area very bad compared to the older Garmin Maps provided by HERE. I've submitted many corrections to OpenStreet. Some they've made, others they have not.

njkayaker

Many drivers don't really look at the map. So, they don't see the problems that show up.

The navigation works better (is more reliable) if one keeps an eye on the map. That's more true for cycling but even true for driving.

njkayaker

It seems you aren't quite getting what I'm talking about (that could be me but it's also not hard to describe).

* Different sites won't produce the same series of track points but, as long as the shape of the track is correct, that shouldn't matter.

* Different maps produce different tracks. That can matter but that's really a "difference in maps" issue.

This is a valid point. Since the route planners basically produce the same file, people should use what is easiest for them to use.

That wasn't what I was talking about. The data produced by different sites will be different. Though, as long as the same map is being used, those differences won't generally matter at all.

You talked about trying different files in RWGPS (the same site) like it could matter. There won't be any difference between using a TCX or GPS track file (the "GPX route" produced by RWGPS file has very different data and won't work at all on the Edges, or most cycling GPS's).

I was talking about the files.

The course files just contain a list of coordinate points ordered by time. There are enough points to accurately follow the curves and turns of the path you intend to follow.

The Garmins provide two modes of navigation: 1- track following/course points and 2- what Garmin calls "turn guidance".

1- With "track following", no maps are used. This method just follows along the track you've loaded to the unit. This method is what the Garmin units that don't have maps use.

2- With "turn guidance", the Garmin generates a second route by walking the track and seeing what roads on the installed map appears to follow. Since this uses the map on the Garmin, things tend to work better if you use the same map to create the course (in practice, the process generally handles differences).

The Garmins with maps (*) let you use both methods at the same time.

(* Except the Touring. Not sure about the "Explorer" models.)

Openstreetmap has features that the Garmin City Navigator maps don't have (like cycleways).

I wonder how you are "submitting" these. I make changes directly to OSM data. None of my changes where "not made".

Koyote

Wow. I went away for an hour, and look at all of these posts!

I appreciate all of the time and effort, and will read these posts more carefully again later. But I do understand things a bit better now.

As I mentioned, Lezyne's own system worked horribly -- I could build the route and load it, and the computer's T-b-T navigation was a complete failure. ridewithgps worked better, but still with errors. I will keep playing with that, and perhaps other systems.

Thanks again, and I look forward to any more info that comes into the thread.

pdlamb

[QUOTE=Iride01;20306679]
Just a guess, but it's really, really hard to pin the same point twice, either in the same site or in different sites. Some of these web sites probably have smarter locate-and-route programs behind them; that could easily be the difference between your experiences of the various interfaces. Some, you just gotta hit within a city block; another might think you're on the east sidewalk going west, so you need to go to the end of the block and cross the street.

noglider 

I should relate what happened to me last year. I made a route for myself using google maps, in unfamiliar territory in the Catskill Mountains of NY. I noticed I had missed a turnoff, so I went back, and it was hard to find. I think I doubled back a couple of times. The turnoff was a dirt trail with a chain and sign that said private property. And if it didn't have those barriers, it was still baffling, because the trail was even steeper than the road which was already killer steep. I was in an area without cell coverage, and I was using my phone, the ridewithgps app, and the route. The display showed me where to go but no map, since I was out of the coverage zone. It was a little scary to improvise, since the roads were very lonely, and I wasn't familiar with the area. It drove home the point that carrying a paper map is still a good idea.

Anyway, I found my way before too long. I had mapped out a 98-mile super hilly route, and at the end of the day, I had ridden 104 miles.

bogydave

Only had a Garmin 520, 1.25 years.
it works good for me , not glitchy.

When I have issues
it usually turns out to be "Operator error".

Mostly just:
The "Learning curve"

njkayaker

You aren't going to pin the same exact point on different sites (or even on the same site). But there's no indication that those slight differences matter at all.

The differences that do matter (mostly due to map data) are almost always obvious (they aren't mysterious).

With "divided highways", they can look like one line at low zoom levels. It's easy to click such that the routing program "snaps to" the wrong side of the street. The same sort of thing can happen if there's a cycleway running right next to a road. If the router "snaps to" the wrong way, it might generate funny routing because it can only cross over between ways at intersections.

It seems that's the thing you are describing.

njkayaker

With Google and RWGPS, you can use satellite and street view to check out odd places/turns/roads.

I would never rely on having cell service.

You can download maps to your phone. Use Galileo, Maps.me (free), or OsmAnd (or others). These programs also let you load the route.

The RWGPS app lets you download the map for a ride too.

I like the products RWGPS provides but I probably wouldn't rely on their navigation app on this sort of ride (I might rely on it in a place I am familiar with).

noglider 

@njkayaker, that's all hindsight. It was my first time doing something like that. Lessons learned.

The ride was for preparation for the Hillier Than Thou ride, which I managed to finish without dying. I later learned that on my Catskills ride, I went on NY State's highest through-road. I won't lie; it was hard.

njkayaker

I'd rate my Edge 800 as "works very well" for navigation. But I also know how it works and how to use it. It's not perfect but it's easy to work through most issues.

(Garmin can have difficulties working out bugs in new units. That's one reason I'd be reluctant get another unit.)

njkayaker

Yes (not surprising). It's just such an essential point that I want to make it clear for anybody else reading this thread.

I started out with "electronic navigation" by using a route loaded to a smartphone app as a backup to using cuesheets.

I now use a Garmin but I'll still often load the routes to a smartphone app. It's hard to review maps on the tiny Garmin screen.

This is where the "lessons learned" might fall a bit short.

One can use a smartphone instead of having to deal with getting and remembering to bring a paper map. Most people always have the smartphone. Paper maps are kind of an exception (and one people might discount as being too inconvenient). Downloading maps is kind of a routine thing for me (after doing it once or twice, it's fast and easy to do and deal with last minute).

Been there. Done that. Got the T-shirt.

Seattle Forrest

There's a lot of land out here, mostly in the mountains, where there's no cell coverage and traffic is rare. That's why I like having maps in the unit. Paper maps are good, too, but a gust of wind got my cue sheet once, and now I really like the electronic version.

Metieval

edge 520 turn by turn is a day late and a dollar short.. always! sometimes a week late.

noglider 

Yes, that's where I was, in the land of a lot of land. It was quite something being in where traffic was so light. It felt so remote!

canklecat

I finally got useful results from Google maps voice navigation this weekend, for the first time. I knew my way home so I wanted to compare the route I'd usually take with whatever Google suggested. Turns out the routes were nearly identical.

Best guess as to why it worked this time? Faster phone and data connection, not relying solely on GPS and offline maps.

My previous efforts Included an older iPhone 4s, and two more recent Android phones: a cheap ZTE Majesty Pro with bare minimally adequate specs, and a decent Moto E4. The Moto E4 is by far the faster of the three.

I've also compared between three prepaid service: Verizon, Simple Mobile and FreedomPop.

• Verizon is its own network. 4G LTE speed is good, usually at least 10 MB/s up and down, often faster, consistently day and night.
• Simple Mobile is an MVNO that leases from T-Mobile. While it claims to have high speed 4G LTE the actual speed is often extremely slow, as slow as 1 MB/s up and down. The only time it's near 10 MB/s or faster is after midnight. Clearly T-Mobile throttles data for MVNOs leasing from them, so you get what you pay for with cheaper services leasing from T-Mobile. No surprise since the T-Mobile CEO has long been openly critical of customers who try to get more for less.
• FreedomPop is interesting. It's an MVNO that leases from AT&T. On my Moto E4 data is very fast, sometimes up to 45 MB/s downstream and 20 MB/s or better upstream (much slower on my iPhone 4s, around 6 MB/s up/down stream, which presumably can't really utilize 4G LTE despite the icon on the screen). But FreedomPop uses VOIP for voice/messaging, so you must use their apps for those services -- the phone's native dialer/messaging are bypassed. But it seems to use data conventionally.

The combination of the reasonably fast Moto E4 and AT&T via FreedomPop made for reliable voice navigation. It updated promptly when I deliberately took different turns from those suggested by Google and got me back on track. At one point I was briefly lost (a neighborhood with many diagonal streets, hard to get oriented at night with no reference points) and the voice updated quickly to get me back on course.

Presumably the iPhone 4s is too slow to work reliably with turn by turn voice navigation, even with a relatively fast data connection.

The ZTE Majesty Pro on Simple Mobile is nearly useless. Slow processor,only 8 GB storage -- most of which is used by pre-installed apps that cannot be removed -- and Simple Mobile's heavily throttled data. It's a very basic phone for voice, texting and limited online access. I shouldn't complain for $10. But I can complain about the Simple Mobile service since it claims high speed 4G LTE and rarely delivers until after midnight. And their customer service is sketchy. They mean well, but appear to be located in Mexico and I often encounter language barriers. Instead of texting or emailing relevant notices, they'll call and leave voice messages which are mumbled and incomprehensible. I tried it for two months and won't renew.

I'll need to try the Moto E4 and FreedomPop/AT&T combo again with a different map and voice navigation app to compare. In the past I've tried maps.me and others with voice navigation and those failed badly -- but that may have been the fault of the older, slower iPhone 4s, even with the fast AT&T data via FreedomPop, because apparently the older iPhone can't handle true 4G LTE. I've seen some older articles claiming the iPhone 4s sold by AT&T with their contract service could handle LTE data, but my phone came from Verizon and while it's compatible with AT&T, it's via FreedomPop, not directly with AT&T. There may be some differences, I don't know.

Anyway, this scraps my plans to keep cheap or older phones on the handlebar for turn by turn voice navigation. It appears older and low end Android phones can't handle this chore reliably. And the Moto E4 is cheap enough, possibly the best value right now in an Android phone.

Incidentally, this is important to me for a few reasons.

Mostly, because I have no natural sense of direction. I have a hole in my head where most folks have a compass. I get lost in city where I've lived for 40 years. Hard to believe I aced a nighttime orienteering course in the military. Probably just got lucky that night.

Also because I know some routes are dangerous. I need to be able to divert to an alternate route quickly.

And a couple of times I've gotten lost when riding with friends visiting from out of town. The last time was after midnight on a cold, rainy night. Kinda tested the patience of my poor friend who was exhausted, cold and miserable.

njkayaker

Google does all the route determination on their servers. It only uses GPS to know where you are (the start of the route). It doesn't use offline maps for route calculation*.

The GPS is required for the app to know if you are following the route or have gone off course.

A faster data connection might help (as well as a faster phone).

If the quality of the route is better, it's likely because Google has better data (and it might be improving the method it uses to choose routes for bicycling).

* Google maps provides the option to download (cache) the map covering your route. That's a relatively recent feature. It doesn't have the ability to calculate an alternative route if you go off track (they might have changed this recently to calculate alternative routes off line).

===============================

Many cyclists want to follow a preplanned track. Google can't really do that.

It's fairly common to not have cell network connectivity. So, it would make sense to have some app that can work for navigation off line.

canklecat

Yup. To minimize variables I've run all three phones simultaneously, each on Google maps. They didn't produce the same results. I'll repeat this test with other maps and voice navigation apps. I've tried maps.me offline on bus rides with both my iPhone 4s and Moto E4 and the Moto E4 got consistently better results -- it responded quicker, adapting to the actual bus headings whenever they varied from the route chosen by maps.me when I plugged in only starting and ending points. The iPhone lagged or missed navigation points. But I haven't tried this test on walks and bike rides.

I know many folks locally who use Map My Ride and/or Ride With GPS. I've tried them but I'm more interested in flexibility, being able to adapt to conditions. Often I ride rural routes and places where construction is underway for housing and commercial development. Occasionally actual conditions don't match any maps. So it's helpful to have an aid for alternate routes.

noglider 

@canklecat canklecat, that's good to read. I left AT&T and tried T-Mobile and some MVNOs, and those experiences were bad, so I came back to AT&T. Saving money was a bad plan.

I get reliably very good results with my iphone 6S and AT&T. So you get what you pay for. It works so well that I have gotten dumber at navigating, and my sense of direction, originally very good, has gotten worse.