Accuracy of POI Locations

 

In an effort to get more familiar with my Garmin gps and practice some tips learned on this forum, I created a short 20 mile route and proceeded to see what campground POI's showed up along my route. What I discovered is that looking for POIs ALONG A ROUTE My Garmin 750 only shows those within 1 mile of the road.

HOWEVER, I also discovered that a large state campground within a mile was missing. I found it's POI located some 17 miles SW of where it should have been. And nearby was another POI that was about 50 miles west of where it should have been. Not too good a confidence builder for my first attempt at using data from POI Factory.

I know most of these POI's are entred by volunteers trying to help, but can't help wonder, since they are volunteers, how many data mistakes are made by people being tired, interrupted by kids, cell phones, etc., etc. I'm sure there is a process to notify someone of errors, but also wonder how often these errors occur and if people lose confidence in the POI databases after a couple mis-directed destinations.

I may never have noticed the misplaced POI's with my tiny screen gps. But after importing the csv file to Streets and Trips, it was fairly easy to spot the well known campgrounds in the local area. The good news is I only saw one missing from about 10 that appeared to be about where they should be.

I have 2 days experience with POI FACTORY, but have discovered 2 significant POI location errors in my first attempt at using one of the databases. My question is, ARE MISPLACED POI's COMMON? Can the databases be used with reasonable confidence for someone wanting to find a campground where expected after a tiring day of travel?

Page 1>>

Reagan said trust but verify

Accuracy should be taken with a grain of salt, the majority of the contributors here are well intended folks and I hope that when they create a file is to help others and not mislead them, but humans sometime don’t have the best of intentions and who knows.

Another problem you may encounter is that when geocoding an address it may not be 100% accurate and many times the creator of the file is not familiar with the area for which the file is being created, to catch those mistakes we all depend on each other to point out where the mistakes are in order to correct them for future users.

So did you report to the POI creator the mistake you found giving the correct coordinates?

In your case I would have verified using other means before departing home the campground locations in your intended route , remember your Garmin is a tool to help you navigate but don’t always follow it blindly and wind up on the train tracks like some other folks have.

--
Garmin 38 - Magellan Gold - Garmin Yellow eTrex - Nuvi 260 - Nuvi 2460LMT - Google Nexus 7 - Toyota Entune NAV

Thanks for the post

I understand your concern that there may be misplaced POIs.

What would be helpful would be for you to post details about the two campgrounds that you are questioning.

Since there are 152 campground files on our site, it would be helpful to know which one(s) you used.

Perhaps we need to discuss how the coordinates of POI locations are determined. I would think that most maintainers set the coordinates at the "along a route" entrance to the POI. This is because if the POI file is to provide an "alert" to someone who has asked for such an "alert" while "along a route", the coordinates must be within 30 meters of the center of the route. (for more information please see this thread http://www.poi-factory.com/node/34321)

With some more information from you, maybe we can find out what has happened.

Yes

Strangely enough, this issue you discovered with campgrounds seems to also apply to Google Maps. Recently, I was looking for Portrait Rock in Massachusetts, and Google's location for it was somewhere along the line of a mile to two miles away from the actual location.

Misplaced POI locations are surprisingly common, and can occur anywhere, even within the built-in POIs on the GPS. However, 17 and 50 miles away from their actual location is a bit much. In those cases, it could be something as simple as two digits in the coordinates switched around when the data was entered. The source data could also be incorrect and the POI creator did not check the data before submitting the file. In some cases (cf. Zip Codes, Canada Postal Codes) the POIs are not intended to be routed to.

With some exceptions* I don't generally download files from this site so I can't tell you whether the files are truly accurate or not. I don't even guarantee the accuracy of my own files, since the world changes and POI files are but a snapshot in time. I try, and I consider my files to be exceptionally accurate, but I will never achieve 100% accuracy. I know this, which is why I encourage the users of my file to contact me if something is wrong.

*Red Light Cameras, Speed Cameras, Rest Areas, and of course my own files.

--
"Anyone who is capable of getting themselves made President should on no account be allowed to do the job." --Douglas Adams

POI accuracy.

I have found more incorrect and inaccurate POI listings on the built-in Garmin POI's than I have on the ones I use from POI Factory. As was mentioned before, there are some files on the Factory that have not been properly verified and need some work. But many of the files here are as accurate and complete as you will find anywhere, thanks to the time and effort put in by the members here. When you find errors, please contact the file maintainer so the corrections can be made.

--
Alan - Android Auto, DriveLuxe 51LMT-S, DriveLuxe 50LMTHD, Nuvi 3597LMTHD, Oregon 550T, Nuvi 855, Nuvi 755T, Lowrance Endura Sierra, Bosch Nyon

Help us help you

All poi files won't be 100% accurate, but what you found seems to be way out of line. I know on the files I make I try and verify using a few different sources. If those sources are close then its ok. If off then I try and find out where the location is. Sometimes that is impossible.
We have found from experience that a company's web site is inaccurate for there locations. That is a mass geocode does not work. A mass geocode is way off. Some maintainers take what they see as correct when in actuality they aren't.

As stated, get us the files you were using and we can look at them. It could be a file that has not been kept up to date.

And as already stated, I also welcome being notified of any errors so I can correct the file.

--
Nuvi 2460LMT.

Another reason

Another reason is the original address used for the geocoding. The address for a street is not always the same from street to street.

A lot of POI's are created in batch to get the cooridinates. If the adddress does not point to the correct place to begin with then the POI will be wrong. The two that I made and one I took over I verified to the best of my abilities. I live in GA and I'm not about to drive all over Ohio or GA to physically verify each address. So using Google earth and POI editor I looked for the correct building and put the flag at the entrance to the parking lot. In this way it will be more accurate. Some places it is hard unless you are familiar with the area.

Example I have seen Lowes and Homdepot both have bad location's on Google. If it is bad on Google when it is geocoded to make a customed POI it will be wrong also until it is corrected for the custome file.

Also hate to say it but there are some files that are just created from a companies address list and never verified. Common problem.

Intrigued

Yes there are going to be mistakes, but as users when we find them, we need to post them and correct them so that other users will not be affected. I admire your ability to bring up this topic as it is a very good one and we need to be careful how accurate anything is when we download it no matter what site it might be even commercial products!

This happens a lot

wknight40 wrote:

Also hate to say it but there are some files that are just created from a companies address list and never verified. Common problem.

A case in point is the Tim Hortons file. We were all looking for an up date file on this site because it was outdated. The file was on Poifriend.com before it shut down. Poifriend was working with Tim Hortons (Timmies gave the info directly to Poifriend. That file is riddled with duplicates and I found about 10 locations in my area that are off on coordinates.
We now have a maintainer for the file (which is great), but it needs fixing.

--
Nuvi 2460LMT.

I live about 30 miles off of

I live about 30 miles off of the nearest Interstate, and have found that a lot of the POIs are listed at the interstate exit. I allways just assumed this was done for travelers so they would get the alert. Many of the alerts I get at the interstate exit are for things 15+ miles from the interstate. Should I be reporting these for better accuracy?

The details to my original post

Well, I wasn't going to broadcast whose POI's were inaccurate, but since several of those who seem to "own" databases have asked, here's the detail:

Database 14045 Combined US Campgrounds by rivopom.csv

Stockton State Park (on Stockton Lake, Stockton Mo)
POI location (wrong) 37.31, -93.581
POI location (correct) 37.60570, -93.75147
Dadeville, Mo (Stockton, Mo?)

RV Express RV Park
POI location (wrong) 37.34205, -93.92903
POI location (correct) 37.34255, -92.93031
Marshfield, Mo.

It would not suprise me if there are very many errors such as these in the POI dadabases. Just the various methods used to record lat/lon is in itself very confusing to many, let alone all the variables mentioned in the above replies to this post, and the fact that untrained people are doing the data input..

It's pretty evident looking at the coordinates above that numbers got transposed in one and who knows what happened in the other.

But my question was, how often does this occur? If I get mis routed and cannot easily find my destination 1 time in 50 I would consider the POI databases valuable and useful. If I get mis routed 1 time in 15, I would rather not use them for my travel purposes.

A previous thread

concerning the same subject with interesting comments and suggestions.

http://www.poi-factory.com/node/26344

--
Nüvi 255WT with nüMaps Lifetime North America born on 602117815 / Nüvi 3597LMTHD born on 805972514 / I love Friday’s except when I’m on holidays ~ canuk

We Understand

Slidein wrote:

Well, I wasn't going to broadcast whose POI's were inaccurate, but since several of those who seem to "own" databases have asked, here's the detail:

*snip*

That's understandable, since it can be embarrassing when it's an obvious gaffe. While not always possible, you want to notify the file maintainer through their contact page if you find something wrong. If you can't get a hold of them that way, then post what you find publicly. The people here are generally appreciative when notified of such things however, regardless of whether it was through their contact page or in a post on the forum.

--
"Anyone who is capable of getting themselves made President should on no account be allowed to do the job." --Douglas Adams

Verifying is not easy work!

First let me say that there are some very dedicated POI maintainers on this site, that take pride in their files and do a lot of work to get the most accurate location they can.

This is not a brag, but I took 10 months to verify the Bank of America POI file. I worked on that file almost everyday after work and on weekends. I thought I could get it done in a couple of months - LOL! See the BofA POI discussion file for some of my whining!

First obstacle was getting a current list of locations. That list is apparently made of un-obtanium. There is not a "complete" list on their website. It is now one of those "search" for locations. You know - type in an address or Zip Code and it returns the 50 closest locations. No responses from BofA to numerous email requests and the local branch here says that I'll have to get the list from customer service. Again - LOL.

I too have been disappointed in some POI files that I have used. Some maintainers make corrections as soon as they can and others don't respond right away. Some never respond - those files really need a volunteer to take a step forward and to raise their hand. I did.

Note: if you do find an error in any of my files - let me know - PLEASE.

--
Metricman DriveSmart 76 Williamsburg, VA

Holy S#!^

With over 10,000 locations, no wonder it took you ten months. smile

--
"Anyone who is capable of getting themselves made President should on no account be allowed to do the job." --Douglas Adams

Thanks metricman

Strephon_Alkhalikoi wrote:

With over 10,000 locations, no wonder it took you ten months. smile

Error Rate

Slidein wrote:

It's pretty evident looking at the coordinates above that numbers got transposed in one and who knows what happened in the other.

But my question was, how often does this occur? If I get mis routed and cannot easily find my destination 1 time in 50 I would consider the POI databases valuable and useful. If I get mis routed 1 time in 15, I would rather not use them for my travel purposes.

Two out of 14045 entries would be a pretty low error rate (if that is all that it is).

I suspect that there there may be less than two errors in metricman's BofA file because he spent so much time on it.

As several have pointed out, google has errors; the built in POIs have errors. This will continue until the companies themselves put out verified coordinates for others to use.

750 branches to close

metricman wrote:

First let me say that there are some very dedicated POI maintainers on this site, that take pride in their files and do a lot of work to get the most accurate location they can.

This is not a brag, but I took 10 months to verify the Bank of America POI file. I worked on that file almost everyday after work and on weekends. I thought I could get it done in a couple of months - LOL! See the BofA POI discussion file for some of my whining!

First obstacle was getting a current list of locations. That list is apparently made of un-obtanium. There is not a "complete" list on their website. It is now one of those "search" for locations. You know - type in an address or Zip Code and it returns the 50 closest locations. No responses from BofA to numerous email requests and the local branch here says that I'll have to get the list from customer service. Again - LOL.

I too have been disappointed in some POI files that I have used. Some maintainers make corrections as soon as they can and others don't respond right away. Some never respond - those files really need a volunteer to take a step forward and to raise their hand. I did.

Note: if you do find an error in any of my files - let me know - PLEASE.

750 branches to close, this shouldn't take to long to fix! I'm glad your one of those that make corrections as soon as they can. I don't want to go to a bank that won't be there. If I find an error down the road I will let you know.

Bank of America (NYSE: BAC) , earlier this week, outlined basic plans to pair down its annual expenses and shed 30,000 jobs in coming years. Some 750 branches nationwide could close, as per comments made during an analysts’ conference call last month, but bank officials have not provided any detailed branch closure information.

--
Using Android Based GPS.The above post and my sig reflects my own opinions, expressed for the purpose of informing or inspiring, not commanding. Naturally, you are free to reject or embrace whatever you read.

After looking around...

jgermann wrote:
Slidein wrote:

It's pretty evident looking at the coordinates above that numbers got transposed in one and who knows what happened in the other.

But my question was, how often does this occur? If I get mis routed and cannot easily find my destination 1 time in 50 I would consider the POI databases valuable and useful. If I get mis routed 1 time in 15, I would rather not use them for my travel purposes.

Two out of 14045 entries would be a pretty low error rate (if that is all that it is).

I suspect that there there may be less than two errors in metricman's BofA file because he spent so much time on it.

As several have pointed out, google has errors; the built in POIs have errors. This will continue until the companies themselves put out verified coordinates for others to use.

If it were 2 out of thousands, that would be wonderful, but I found 2 on my first look at less than 30 POI's in my local 50 mile area.

Now that I've had a little time to look around on POI I may be getting a different picture of how this all works. My initial impression was that anyone (ie Bubba the camper) could email in some numbers and a name to POI Factory and someone would stick them in a campground database.

It now occurs to me that the over 5000 databases are built and maintained by maybe 2000 owners. The implication being that at least these folks have some knowledge of spreadsheets, database structure and probably do their own input and editing. If that's true it eliminates multiple parties involved with data manipulation. I'm not sure what POI Factory's involvment is other than to maintain the website. Do they set standards for those making the files available? Does anyone monitor complaints about accuracy, etc., and counsel or remove those db owners who have a difficult time?

I'm still waiting to see if anyone is brave enough to guesstimate what the overall error rate is for destinations that aren't found in a reasonable time.

We Help Each Other

The POI creators happily accept input from others regarding things to add to their files. The thread that canuk linked to shows an example of this, as in it I mention working on my first POI file (Historic Vessels of the United States and Canada) and I had some three different people provide me with assistance. Granted, most of the assistance was redundant has I had already found those particular ships, but at least one was useful.

Back when I first wrote Historic Vessels and for about a month afterward, it was in CSV format, written using notepad. The file could be opened in Excel of course, but for such a simple file notepad was more than sufficient. When I changed formats to GPX, Excel wouldn't work since GPX is an XML-based schema. While I could have used Notepad++ to hand code the file, Extra POI Editor by our friend Turboccc was a much better choice since it was designed specifically for the GPX format.

Little knowledge of spreadsheets or XML schemas are required in creating POI files. Notepad is sufficient for most CSV files, and Extra POI Editor is best for a GPX file. We do our own input and editing, on our own time, and often can spend upwards of ten hours on some days working on the file, just to make it as good as it can be. POI Factory's involvement is limited to checking CSV files for formatting errors (it doesn't check GPX files at all) and hosting the files.

Complaints about accuracy need to be taken up with the individual who created the file, and there is no "counseling" or "removing" of POI creators. The reason this site works is not because of rules, but because everyone tries to help each other out. We all have our disagreements of course, but in general we watch each other's back.

--
"Anyone who is capable of getting themselves made President should on no account be allowed to do the job." --Douglas Adams

Once, I was directed to a

Once, I was directed to a dark rural street when my GPS told me there was an A&W there. That was my first memorable GPS goof!

--
~Jim~ Nuvi-660, & Nuvi-680

LOL

Slidein wrote:

Does anyone monitor complaints about accuracy, etc., and counsel or remove those db owners who have a difficult time?


POI Factory my name Peggy you got problem?
Peggy? Third time I call its time I speak with a Supervisor
Supervisor is genius I transfer
Transfer!
Transfer!
Transfer!
Transfer!
Transfer!
Hello my name is
Peggy
Come ON!!!
Hello?

--
Garmin 38 - Magellan Gold - Garmin Yellow eTrex - Nuvi 260 - Nuvi 2460LMT - Google Nexus 7 - Toyota Entune NAV

Agreed

Strephon_Alkhalikoi wrote:

The reason this site works is not because of rules, but because everyone tries to help each other out. We all have our disagreements of course, but in general we watch each other's back.

annoying

It gets a little annoying when a newbie comes on and the fefore they even know the site they start complaining and chalenging what is poste and how it is done. I see everday where one or more POI maintaineers are updating their POI. I wish people would do a litte research and read before commenting on things. sad

--
johnm405 660 & MSS&T

Agreed

That is why either this video or this video should be noted as a faq. Personally, I prefer the first one. :)

--
"Anyone who is capable of getting themselves made President should on no account be allowed to do the job." --Douglas Adams

agree

Strephon_Alkhalikoi wrote:

That is why either this video or this video should be noted as a faq. Personally, I prefer the first one. smile

I like the first one also it should be required watching when first getting on the site.

--
johnm405 660 & MSS&T

Complaint Department

Looking back at the last three years of being a maintainer I can’t ever remember receiving a complaint, some have written me to correct locations with information they provided and others to remind me some of the files need a little refreshing since they know of places that have either open or closed (something a maintainer may never be aware).

I have contacted others here to fine tune coordinates of places I’m familiar with but never to complain.

There is definitely room for improvement on some of the files we maintain; and as a user you also have responsibilities to check some of the files before embarking or running the risk of wasted time and fuel driving to erroneous locations.

Here is the rule I go by when using files made by me and other members.

MrKenFL does an excellent job maintaining the Rest Areas Combined file, he is very conscious of his work and I trust the info, but if he messes up on one it is no big deal, just hold your bladder and drive a little further to the next Rest Area.

JM Red Light Camera file is very good but doesn’t mean you get a license to run the light, do so at your own risk since he may not have 100% coverage of all 57 states [sic] and Canada.

Restaurants files (including my own) if it hasn’t been updated since 2007 call ahead, restaurants open and go out of business faster than you can burn a toast.

National Monuments, Parks, Hospitals, Police and such are for the most part accurate regardless of when the file was made, can’t remember the last time they moved Yellowstone.

Remember when using a GPSr don’t follow it blindly, regardless if the POI came with the unit or is one you downloaded on the Internet.

Now that we got that out of the way who wants to run the complaint department or is that JM’s job?

--
Garmin 38 - Magellan Gold - Garmin Yellow eTrex - Nuvi 260 - Nuvi 2460LMT - Google Nexus 7 - Toyota Entune NAV

Shoot the Messenger

Some of you seem to be off on a "Shoot the Messenger" crusade.

If you bothered to read my first post it said that on my first attempt at using a POI file I found 2 POI's that were significantly mis-located. Someone suggested 2 out of 14k is not bad. Later I mentioned that those 2 were out of about 30 POI's in my local area. 2 out of 30 is NOT an error rate I feel comfortable with and what I asked was, whether misplaced POI's were common. The next 10 replies or so went on to state how difficult it was to accurately locate POI's due to a whole myriad of variables.

Now that I understand better how poi files are generated, it seems EXTREMELY difficult (if not impossible) for someone to sit down and create an accurate data base of thousands (in the file I was using) of POI's throughout the US, by copying information primarily gleaned from various internet sources.

I came here to see if a POI tool could be useful for a cross country trip I am about to take in my camper. I was looking for something that was at least as accurate as Garmin (what little they have) or the Escapees campground poi's. After finding 2 errors in 30 on my first look here, I was disappointed. Was that an anomale? Was that "normal" considering the complexity? No one seems willing to venture an opinion.

I will take the POI file with me and see how it works (as well as a couple other poi files). Traveling in a Truck Camper, it's not the end of the world if my destination isn't found, but it's certainly more comforting if it is.

I hate to bring this up since my comments about standards, inspecting the process or validating poi files seemed to rile some, but it seems to me that one way to reduce the number of poi errors would be to have POI Factory members volunteer to be "Validators". For small files this probably is not necessary, but for those huge national files, members might be asked if they would validate POI's within 50 miles of their location. Give a special icon for an incentive, etc.

.

Slidein wrote:

I came here to see if a POI tool could be useful for a cross country trip I am about to take in my camper. I was looking for something that was at least as accurate as Garmin (what little they have) or the Escapees campground poi's. After finding 2 errors in 30 on my first look here, I was disappointed. Was that an anomale? Was that "normal" considering the complexity? No one seems willing to venture an opinion.

It's apparent from your comment about Garmin's built in POIs that you are a gps novice. You will quickly learn that the built in Garmin POIs are woefully inaccurate. That's one of the main reasons this site exists.

All depends on the maintainer

Slidein wrote:

I will take the POI file with me and see how it works (as well as a couple other poi files). Traveling in a Truck Camper, it's not the end of the world if my destination isn't found, but it's certainly more comforting if it is.

Like yourself, I like programming lots of potential waypoints into my Nuvi ahead of a road trip. The software tools we have allow taking a satellite view and I do so, with every waypoint I put in there.

I don't use a lot of POI files due to the same inaccuracies that you have mentioned. There were too many folks who were just batch geocoding address files. Many others were actually checking their entries with satellite view for accuracy. But there is no way to tell for sure. It only takes one bad or bogus POI to really sour a road trip. An independent verification or spot check would add to the confidence level.

POI Accuracy

The biggest headache I enctyounter whenm looking for some POI's is when they are located in a large shopping mall area or strip mall with a large parking lot.

Usually the GPS will take you to the main entrance and it can take a long while to find the actgual location of the restaurant/eatery you are looking for.

I have found many food POI's that when I arrive, are closed or out of business.

yes, some have thin skins

Slidein wrote:

Some of you seem to be off on a "Shoot the Messenger" crusade.

I did read the post in its entirety and understood your concern.

Slidein wrote:

If you bothered to read my first post it said that on my first attempt at using a POI file I found 2 POI's that were significantly mis-located. Someone suggested 2 out of 14k is not bad. Later I mentioned that those 2 were out of about 30 POI's in my local area. 2 out of 30 is NOT an error rate I feel comfortable with and what I asked was, whether misplaced POI's were common. The next 10 replies or so went on to state how difficult it was to accurately locate POI's due to a whole myriad of variables.

As has been stated, there are a myriad of reasons errors like this are found, EDIT: (using a very big brush) the primary two being the software used to find the coordinates and the failure of the maintainer to verify the coordinates are correct. Now, I'm not going off on maintainers not checking their files, the job can be tremendous with large files. Metricman stated it took him 10 months to compile the BofA ATM file and he knows there are some errors still in the file.

Not excusing either, but as an explanation probably the most common geocoder used is Google's. On average I find about a 30 to 40% error rate from the returned results. That sounds high, but it also depends on what precision you are requesting. For stores and places of business, you normally want it to be "spot on." Google will more than likely get you within a 100 feet or closer. We have no control over that other than to look at each result returned and fine tune it to the actual location. If Google can't resolve an address to a street level it will back out till it does find something it knows where it is. That can be the geographic center of a zip code, city, or even a state. This is where it becomes necessary for the maintainer to step in and correct those errors.

Slidein wrote:

Now that I understand better how poi files are generated, it seems EXTREMELY difficult (if not impossible) for someone to sit down and create an accurate data base of thousands (in the file I was using) of POI's throughout the US, by copying information primarily gleaned from various internet sources.

Thank you for admitting that building a correct POI list can, and often is, a difficult task.

Slidein wrote:

I came here to see if a POI tool could be useful for a cross country trip I am about to take in my camper. I was looking for something that was at least as accurate as Garmin (what little they have) or the Escapees campground poi's. After finding 2 errors in 30 on my first look here, I was disappointed. Was that an anomale? Was that "normal" considering the complexity? No one seems willing to venture an opinion.

I will take the POI file with me and see how it works (as well as a couple other poi files). Traveling in a Truck Camper, it's not the end of the world if my destination isn't found, but it's certainly more comforting if it is.

I hate to bring this up since my comments about standards, inspecting the process or validating poi files seemed to rile some, but it seems to me that one way to reduce the number of poi errors would be to have POI Factory members volunteer to be "Validators". For small files this probably is not necessary, but for those huge national files, members might be asked if they would validate POI's within 50 miles of their location. Give a special icon for an incentive, etc.

Jon does have a validation program for Garmin formatted CSV files. When a file is submitted it is run through the checker and the results are posted along with the file. As an example, here's what's posted for one of my files: Includes 5601 locations in the following areas:

United States: AL, AR, AZ, CO, CT, DE, FL, GA, ID, IN, KY, LA, MA, MD, MI, MN, MS, MT, NC, NE, NJ, NM, NY, OH, OK, PA, SC, SD, TN, TX, UT, VA, WA, WI, WV, WY
some may be in: CA, DC, IL, MEX, MO, NV (near a border)

Now, have I personally verified each location is correct - No and I admit I never will as it is a repost and a reformatting of a list published monthly by the company. I state this file is from the company website in the description also. The value I add to the company's file is the creation of a GPX version which allows a bluetooth enabled unit to dial the location.

What many feel is "proper netiquette" would be to send the maintainer a private message using their contact tab unless there is an open forum discussing that particular POI. Corrections are then posted to the forum. Now, some maintainers are no longer active so a note to JM, the webmaster would be in order. it is possible for him to make corrections to a CSV file, but I don't think Jon wants to muck around in a GPX.

Don't loose heart, keep on using both the site and the files taking the time to point out errors you encounter. Just be aware that some won't take kindly to any criticism no matter how well intentioned.

--
Illiterate? Write for free help.

Achievable Accuracy A

Achievable Accuracy

A standard GPS receiver for civil use offers an accuracy down to a few meters. In praxis the number and geometry of the received satellites influences the accuracy considerably, and in daily use, accuracies of about 20 m can be expected. More sophisticated GPS receivers as they can be found for land survey cost several thousand dollars and achieve an accuracy of a few centimetres.

With selective availability activated by the government, receivers achieve accuracies of approximately 100 m (these declarations are true for 95 % of all cases). After deactivation of the selective availability the accuracy rose to approximately 15 m, depending of the number and position of available satellites.

Selective availability was a degradation of the accuracy imposed by the Government for defense security safety. That has been lifted but still the accuracy while improved will not be pin point unless you are working in certain Industries like land surveying and engineering that need much better point accuracy for coordinate system.

Coupled with human errors in making POI files don't expect too much of an accuracy.

--
Garmin Nuvi 260W Garmin Nuvi 1490T If you think knowledge is expensive, try ignorance.

Not shooting the messenger

Slidein wrote:

Some of you seem to be off on a "Shoot the Messenger" crusade.

...

I came here to see if a POI tool could be useful for a cross country trip I am about to take in my camper. I was looking for something that was at least as accurate as Garmin (what little they have) or the Escapees campground poi's. After finding 2 errors in 30 on my first look here, I was disappointed. Was that an anomale? Was that "normal" considering the complexity? No one seems willing to venture an opinion.

I will take the POI file with me and see how it works (as well as a couple other poi files).

...

I hate to bring this up since my comments about standards, inspecting the process or validating poi files seemed to rile some, but it seems to me that one way to reduce the number of poi errors would be to have POI Factory members volunteer to be "Validators". For small files this probably is not necessary, but for those huge national files, members might be asked if they would validate POI's within 50 miles of their location. Give a special icon for an incentive, etc.

@Slidein, I do not think that the comments in this thread were "shooting the messenger". Rather, I think that they were attempts at explaining the difficulties of creating files. Perhaps when you have tried to create some files yourself, you will have a better understanding of the commentary in this thread.

I do not think that it has been mentioned that Garmin is not responsible for the built-in POI files. Garmin gets them from a third party. We have never been able to determine why various individual POI locations are included - and on the other side - why other POI locations are excluded. As an example, my Subaru dealer is not included but has been in business for many years.

As others have already mentioned, many of the built-in POIs are no longer in existence or have moved to a new location. There is no way to update the built-in POIs other than when Garmin decides to put them into a new release. On the other hand, poi-factory files are updated on a more regular basis.

You can contribute by taking the time to give us feedback after your trip. You will be able to check out many of the campground coordinates - regardless of whether you actually overnight there. If 1 out of 15 of them turn out to be grossly misplaced, you will probably conclude that the files here are not what you want. However, we would still want to know because we would want the errors fixed.

Camping files in particular

The camping files in particular need overhaul, imo.

Almost 4 years ago I discovered POI Factory for the same reason as you, looking for camp sites on an upcoming trip. The COE and State Park files that I got here, if not usually exact, were good enough that I found what I was looking for visually.

'Camping' is tough to begin with because everyone has preferences (walk-in, anything tent, allow RV's, and what size RV, don't allow RV's, etc.)

Over those 4 years many have attempted to make the 'perfect' list and rather than start from scratch they have built around various lists either from here or other places. Due to the huge number of locations and the inherent probability of accuracy loss each time someone transcribes existing info, I avoid the 'combination' camping files and prefer making my own from the smaller specialty sets here. Hikers know walk-in sites best, others know tent-car sites best, others RV parks, and so on.

However, anything in a poi file should be considered a navigational aid rather than an absolute, imo. I recently tried to create a poi file from a federally mandated US govt. listing that originated from each federal department responsible for the camping facility (Park Service, BLM, COE) and it was both inaccurate and incomplete enough to be unusable.

But to answer your question about our opinion of the accuracy of the files here.. imo the grossly inaccurate files are in the minority, but they do exist.

One thing I'd certainly advocate on this site is use of gpx files due to their granularity and the ability to create an 'audit trail' of sorts.

The other would be a requirement to list the sources of some of these huge 'combined' files, if for no other reason than to be able to make corrections in all of the files when an error is found in one.

Lastly, with more time on the site I think you'll find that people verifying pois close to them is what happens here better than most sites I've found, given the wide variety of poi 'topics'. Yes, a backpacker's site will probably have better walk-in site accuracy, and a RV forum better files of RV locations, but there is a policy here of not stealing and posting other people's work without explicit permission, so we all do what we can.

--
It's about the Line- If a line can be drawn between the powers granted and the rights retained, it would seem to be the same thing, whether the latter be secured by declaring that they shall not be abridged, or that the former shall not be extended.

@JD4x4

thoughtful comments. Thanks

just wondering...

To all of you people, I would like to know what do you do to get better information for your files..

Since I do the Offbeat poi file, it can sometimes be hard to find the right spot, or to find where to mark the spot. I do try to be right on but it is hard.
Some of my rules are that anything that is in my File must have a address, and if possible a phone number. If I see the mark on google, I will place the marker there and get the reading. Most of my findingins is with ExtraPoi edit and I do use the map to see if I can locate the place. This is hard since I Find alot of great places but I can not really go there. I must count on another way to find this location. If anyone knows a better way to be more Accurate, I would love to hear it.

So how do you find your location and make your poi file accurate?

Thanks for this information..

Bob

--
Bobkz - Garmin Nuvi 3597LMTHD/2455LMT/C530/C580- "Pain Is Fear Leaving The Body - Semper Fidelis"

bobkz

Imo, including the address, and if possible the phone number is the best anyone can ask for (aside from the coords being as accurate as possible!).

TurboCCC is the only one to know how he coded his map & look-ups, but I've programmed the Google API enough to know that they have some additional 'layers' that are generally available on their map site which don't show up in 'plain vanilla' api searches & displays. I've found that address & location corrections will often show up on a web map search before they do in a programmed api query. So, sometimes using the web version of Google Maps might show up something that doesn't show in the EPE map.

--
It's about the Line- If a line can be drawn between the powers granted and the rights retained, it would seem to be the same thing, whether the latter be secured by declaring that they shall not be abridged, or that the former shall not be extended.

Hello all of you I maintain

Hello all of you
I maintain the campground file and work on it every day. If i had to restart the file today i would take a different approach.
I gathered the information from sites that were inacurate even the big sites, street&tripp is inacurate, Google earth is inacurate,Woodhalls is inacurate
Garmin is inacurate i think this Campground file is about the most acurate you will ever get since it is updated regularly,
The main file has been downloaded 6,400 times so people use it. The problem is i rarely get feedback on errors, im on my own which requires 4 to 5 hrs a day to correct
Lately i have had help from a few people and i thank them the file has improuved a lot.
(Slidein) sent me the information error which was corrected immediately People should do the same thing GIVE FEEDBACK.
I am adding accuracy,addresses and phone but this takes time the file will never be 100% precise but if people give feedback on errors il will improuve rapidely

Robert

POI's Need vs Convenience

camping rivopom wrote:

Hello all of you
I maintain the campground file and work on it every day. If i had to restart the file today i would take a different approach.
I gathered the information from sites that were inacurate even the big sites, street&tripp is inacurate, Google earth is inacurate,Woodhalls is inacurate
Garmin is inacurate i think this Campground file is about the most acurate you will ever get since it is updated regularly,
The main file has been downloaded 6,400 times so people use it. The problem is i rarely get feedback on errors, im on my own which requires 4 to 5 hrs a day to correct
Lately i have had help from a few people and i thank them the file has improuved a lot.
(Slidein) sent me the information error which was corrected immediately People should do the same thing GIVE FEEDBACK.
I am adding accuracy,addresses and phone but this takes time the file will never be 100% precise but if people give feedback on errors il will improuve rapidely

Robert

Robert, thanks for your post and updating the Campground file. I'm the guy that started this thread and have only been on this website a couple days, so can't offer much help on HOW to fix things. Your 14k poi file seems like a tremendous undertaking, especially considering the impact if a location is wrong after someone travels off highway very far dragging a trailer.

There seems to be mixed feelings about the importance of POI accuracy. It occurs to me one of the reasons some may feel POI accuracy is less important than others is due to the circumstance. If looking for a Walmart or a bank or fast food retaurant ends at a wrong location, simply roll the window down and ask the first person you see where it is. Or drive another half mile and look for the sign. Satisfaction in this case comes from having your gps guide you correctly to the first parking space at the entrance door. If it doesn't, the impact is minimal, so notify the file owner.

Then consider the traveler looking for a campground in a remote location after a long day's drive. If you end up 2 miles down a gravel road with no campground, is it a minor inconvenience like above?

I think it is important to have CAMPGROUND or other destination POI's that could have significant impact if wrong, be as accurate as possible BEFORE PUBLISHING. This may be only 5% of the total file population. Somehow, additional emphasis needs to be placed on these type POI's, especially if the location is not near a population point, major road, etc.

POI Factory offers a Red Light POI for a fee. With that comes some degree of validation of the data.

"POI Factory's camera files are the best available for USA and Canada. We carefully research reports from the POI Factory user community and are very selective about what gets added and removed."

How do they validate? Could something like this be done with Campground or similar high-impact-if-wrong poi files? If you have 6400 downloads I wonder if you might find 1000 VALIDATORS throughout the US who would volunteer to check out their local areas if they thought there were controls in place that allowed publishing a very accurate Campground map, BEFORE PUBLISHING.

I will be gone for several weeks on a 2000 mile trip. Hope to get time to use your Campground POI's if the weather cooperates.

Kirk

campground file

If you take this project at heart then participate in its developpement send FEEDBACK

Robert

@slidein

While your concerns about going a couple of miles down a dead end road and not finding the campground using GPS directions are valid, I wonder if you would have the same amount of poor locations relying on paper maps and printed directions.

--
Illiterate? Write for free help.

Campground File

Robert,

Gee, thought that was what I was doing to the extent I am able at this point.

Off to trip

Box Car wrote:

While your concerns about going a couple of miles down a dead end road and not finding the campground using GPS directions are valid, I wonder if you would have the same amount of poor locations relying on paper maps and printed directions.

Probably not. Don't remember ever having a CofE, State or Fed map showing campground locations that were not accurate. But that's not the point. Just suggesting there may be ways to make things better here. Hang in there...I'm about idea'd out!

Digital

Slidein wrote:

snip..
Don't remember ever having a CofE, State or Fed map showing campground locations that were not accurate. ..

Providing the digital data is another story, however-
https://www.recdata.gov/RIDBWeb/Controller.jpf
I've already turned this data into gpx, and it's far from usable.

So, we do the best we can & all of us (hopefully) validate as we can.

--
It's about the Line- If a line can be drawn between the powers granted and the rights retained, it would seem to be the same thing, whether the latter be secured by declaring that they shall not be abridged, or that the former shall not be extended.

If the messenger comes to complaint shoot’em

I think where this thread went sour was with Kirk (Slidein) choice of words at least it did with me, had he use the word FEEDBACK as Robert (camping rivopom) suggested instead of COMPLAINT the replies he got would have been less harsh.

Lets face it, campgrounds and rest areas are among the hardest to get right, you are either relying on posted information that someone else gathered or travelers feedback, the other files like restaurants, hotels, shopping are easier because you have an address and ZIP to use for getting the coordinates and in the worst case the POI points to strip mall and not the precise location of the POI

I recently gave FEEDBACK to MrKenFL about some rest areas in TX he didn’t even know existed, I’m willing to bet the same occurs with campgrounds.

--
Garmin 38 - Magellan Gold - Garmin Yellow eTrex - Nuvi 260 - Nuvi 2460LMT - Google Nexus 7 - Toyota Entune NAV

I do the KOA Campground file

I do the KOA Campground file. I consider it to accurate. When I took over the file it was unusable because it was so bad. The info came from their website and their catalog. I did find geocode errors in it that I had to fix. There have been corrections made periodically. The thing I found difficult was looking at a satellite view and trying to determine if I was looking at a campground or a trailer park. I have to rely on users to get me better coordinates.

I understand the concern of traveling down a back road and then not finding a campground. How do you just turn around a big rig?

One thing I build in to my files is a phone number. The number is displayed. I do not like to maintain a csv and a gpx file (it can be converted). I strongly believe a phone number (if available) is paramount in a file.

That is what is missing in some of our files. I use the motel files and have had to create my own just to get a phone number. I want to call ahead to see if there is a vacancy. Not just go there and find no room.
I have suggested I would help on some of the motel/hotel files to add numbers, but it has fallen on death ears (not what I expected).

--
Nuvi 2460LMT.

Phone Numbers Very Important

pwohlrab wrote:

One thing I build in to my files is a phone number. The number is displayed. I do not like to maintain a csv and a gpx file (it can be converted). I strongly believe a phone number (if available) is paramount in a file.

I agree a phone number is very important.

Thanks for taking the extra steps to get them.

A good phone number is absolutely necessary

pwohlrab wrote:

...I strongly believe a phone number (if available) is paramount in a file....

I couldn't agree more. The campground business has become one of the most volatile businesses around in recent years. There is constant churn where one goes out of business completely or changes from an RV park to a trailer park. Sometime the name changes, or last year it might have been a KOA and this year it is an independent without affiliation. Of those campgrounds who have websites, many do not list a good street address partly because their rural locations don't have one. And sometimes the directions are something only a local could follow. For those of us who maintain such files, we know how much work it is and thoroughly appreciate any time someone sends us a note with an update or a correction.

A phone number gives the camper an opportunity to determine if the campground is open, has availability and also if there are some restrictions that may not be suitable for a given rig. It also gives you a chance to verify that your GPS is pointing you in the correct direction to the destination.

--
"There's no substitute for local knowledge" nüvi 750, nüvi 3597

To ALL the maintainers... a

To ALL the maintainers... a million thank yous!!!

--
~Jim~ Nuvi-660, & Nuvi-680

Yours would be tough...

bobkz wrote:

To all of you people, I would like to know what do you do to get better information for your files..

Since I do the Offbeat poi file, it can sometimes be hard to find the right spot, or to find where to mark the spot. I do try to be right on but it is hard.
Some of my rules are that anything that is in my File must have a address, and if possible a phone number. If I see the mark on google, I will place the marker there and get the reading. Most of my findingins is with ExtraPoi edit and I do use the map to see if I can locate the place. This is hard since I Find alot of great places but I can not really go there. I must count on another way to find this location. If anyone knows a better way to be more Accurate, I would love to hear it.

So how do you find your location and make your poi file accurate?

Thanks for this information..

Bob

For me, mine are restaurants and a grocery chain, so it is quite simple. I look for the place that will get you to where you will park your car after driving to the location. If it is the building, I place the mark on it. Sometimes the parking lot, or the entrance to the parking lot, depending on the way it may route you as it tries to turn from the closest road.

For yours, it may not be as simple. For me, the most frustrating part is using my own POI, and then having to struggle a bit to find the location as it is not always accurate.

I don't think there is a perfect solution unless we were able to physically confirm each location, so it is essential to have input from others to make the POIs as accurate as we can.

I agree with the others, many of the POIs listed here are far better than what you can find pre-loaded.

--
Streetpilot C340 Nuvi 2595 LMT
Page 1>>