Google Maps in Thailand

After my rant about our map last year, it’s now time for a new rant: shitty Google Maps.
My old Garmin Oregon is about to fall into pieces (the buttons are broken, I can switch it on with a screwdriver only, GPS becomes erratic in early afternoon or is not available at all for a couple of hours), and I decided to buy a smartphone with GPS. Of course, Android devices come with a lot of things from Google pre-installed, among them Google Maps.
Google Maps allows for offline-use also: you can download some areas before your trip. These offline maps contain roads (incl. routing info) and some POIs. Note that there is no specific bicycle routing available for Thailand, I had to use car routing instead. POIs contain e.g. “restaurants” - without any differentiation between simple noodle soup shops and better restaurants. But POIs do not contain accomodation - ehm, is it really hard to find some food in Thailand’s hinterland, and no problem to find a ressort? My experience is rather the other way round…
Well, of course, you can go online and get that extra information online. With my true move SIM card, Google Maps had an abysmal performance even when the connection was great (typically twice as much download speed as my terrestrial line at home!). But there were also some areas were true move was not available, and on some occasions my German SIM card could not connect to AIS or dtac either (i.e. no network at all available there).

Let’s continue with “resorts”. Last year I said that most of the resorts available are still missing on our map. That seems also true for Google Maps, though Google knows many more resorts than does OSM.
But there is a catch: Google knows some resorts which do not exist. Some cases may be not so detrimental, e.g. when a place was duplicated. In Phangnga, there are “Phang nga inn” and “Phang-nga inn” directly next to each other on Google Maps - it is only one place, and it uses the latter spelling. Worse are places in nowhere which you then find out do not exist at all. They are really shit!

Let’s all work together to ensure OSM is in a better quality here in Thailand than our competitors.

With the map getting more and more complete, the tasks shift from adding new things towards maintaining existing data. This avoids having similar issues than what you observed with Google.

A side note regarding the Oregon buttons. This is a known issue. The rubber buttons can be repaired using something like sugru. Search the web for instructions.

There exist plenty of apps for Android using OSM data, so your favorite map is also available offline.

Ha, yes a lot of Google POI’s were done using algorithms from the Street View images … and it shows.
Sometimes Google does get new places listed quite quickly, but I took great delight in telling them yesterday, to include the complete Chiang Rai-Chiang Saen bypass road (CR.1063), as it been open a month now.
I mentioned the full details, including U-turns could all be found on OSM !! :smiley:

Of course, we should try our best to improve “our” map - we can really do it, in contrast to Google Maps.

But let me continue this rant on Google quality: it is important to know the problems of Google, otherwise someone might try to use Google “to remember things he saw on ground” but may actually be mislead by terribly wrong data!

Look at road names and road numbers. Google “knows” far more of them than we do.
But … what do you think of a road named “Soi 4008”, a minor road connecting to road 4008? To me, that does not look correct (though I did not check on ground).

Reference numbers of DRR roads almost always lack the changwat prefix. Only occasionally, e.g. Krabi 1020, they have a correct number.

Also many roads connecting to a numbered road somehow receive the number from there - take a look at the many roads with number 4194 near Phipun at,99.6011378,15.42z/data=!5m1!1e1

And no let’s come to very bad part of Google Maps. I mentioned many roads with numbers of a nearby highway above, and that’s also true near Huai Nam Sai lake,

Road Ns 4096 was missing completely in OSM, but a milestone showed that it is not a dead end:

Road 4270 was wrong in OSM and Google near the dam.
I decided to check that section of “4270” which leads into the forest in northern direction, just east of the dam:,99.8145675,14.14z
It was missing in OSM.

It started as an asphalted road, but rather low quality. Shortly afterwards, the road was no more paved, but not so bad either:

Steep section had concrete pavement, and steep it was: sometimes beyond 20%.
Eventually I reached the pass height, and went down on the other side. Then I found a school on the left side of the road, and the school roads were even paved with concrete (in the aerial images, you’ll see forest only):

The road deteriorated terribly afterwards, cycling in the soft sand was not easy. When I arrived at an unmanaged ford, I decided to look at the aerial images of Google Maps, but … there is no phone connection available in that jungle.

Hence I returned to the junction where according to Google road 4270 ends. I asked a local man if I could get to Wang Hon on that road to the west, and he confirmed that.
According to Google Maps, road Ns 4096 was 5.2 km away and could be reached by normal car in 13 min - i.e. at an average speed of 25 km/h.

It started off worse than the other road, a very bad smoothness, and sections with concrete lanes (instead of full concrete pavement). Near the top, I missed the junction to road Ns 4096 and got stuck in a rubber plantation. Because the “road” to take leads down very steeply, and is not paved at all. You cannot do that by car, and with a 4WD you ought to have some experience also. I pushed my bike down.

That terrible way is to the left, and can hardly be seen on the photo:
Later on, it got better, and you can play with the images of these jungle roads I uploaded to Mapillary.

So take care when you travel with Google Maps on minor roads!