Garmin vs Smartphone GPS?

[Srini]

How much more accurate are the Garmin GPS units? I've tried my LG Optimus Net not-so-smart phone's GPS for an off road path through heavy tree forests and it's quite inaccurate. Will using a Garmin help or do the presence of trees make all GPS inaccurate? Note: the phone was in my pocket and had a plastic cover on it, and I now found out that having it out in the open is better for accuracy
I need a very accurate means of GPS because I'm trying to map some trails
Thanks and cheers,
-Srini

[berkeleydb]

Finally it will depend on the specifics of each device, but dedicated GPS devices should in general give better results.  (Including things like giving better results even when kept in a bag, etc.)  I do plan to go for a dedicated GPS device myself -- primarily for battery life (#), but I expect the accuracy too to be better.  (# -- ideally i'll prefer a device that runs on AAA or AA batteries, so I can carry spares, and change on the go; but even with one set, I hope I can get ~12hrs or more of run.)

Garmin is one brand; another that a friend of mine told me about recently, is Holux.  

The smartphone GPS can then be used only when you actually need to refer to a map, or when you need to checkin your location to facebook, etc.

-{db}.

[jai.r...@gmail.com]

The Garmin or specialised GPS receiver will be more accurate than smartphones when it comes to giving the location coordinates. The application could add additional inaccuracies (depending on how it maps the coordinates to the map).

That said - it should not matter for your specific activity of mapping trails both should provide trail information which can be discerned by the human eye.

Jai
Sent on my BlackBerry® from Vodafone

--
biking conversations on the world famous "Bangalore Bikers Club" :)

are you a part of the bicycle racing scene?
Visit www.bangalorebicyclechampionships.com for more details

[Srini]

Jai, even I thought the inaccuracies would be too slight to notice! but the problem is, it's off by quite a lot, up to 25 feet! Plus, we are mapping a trail in relation to other paths, so everything must be very distinct.

Application inaccuracies, good point on that! I'll check out what's the most accurate software too rather than only looking on the hardware side!

[deepakvrao]

The Garmins are accurate enough to show you which side of the road you were on when you magnify the map. So, on an out and back, it distinctly shows your separate out and back paths.

[deepakvrao]

To add, ridden thru some pretty thick forests too, and never had a signal loss. Thats on road through forest though, and never on a off road thru forest.

[Sreepathi Pai]

On Sat, Dec 29, 2012 at 11:26 PM, Srini <srinidhi...@gmail.com> wrote:
> Jai, even I thought the inaccuracies would be too slight to notice! but the
> problem is, it's off by quite a lot, up to 25 feet! Plus, we are mapping a

7m is very good accuracy for low-end GPS devices (like the old
Magellan that I have). Good units will also report confidence in their
position (DOP values) along with the position itself which will allow
you to identify issues like tree cover, lack of satellites, etc.

To get more accuracy, look for features like WAAS (which guarantees 6m
accuracy or more when available), and GLOSNASS support like this GPS
unit:

<http://www.amazon.com/Garmin-eTrex-Worldwide-Handheld-Navigator/dp/B00542NVDW/ref=pd_cp_e_2>

You'll need GAGAN (i.e. the Indian WAAS equivalent) to begin operation
for < 3m accuracy in the Indian subcontinent. It is highly unlikely
that handheld units will support that much accuracy cheaply.

Summary: It's unlikely that focusing on Garmin vs Smartphone is going
to you more accuracy, what you have is probably the best you're going
to get in the near term.

--
Sreepathi Pai

[Srini]

I may have misjudged the value of 25 feet or 7 meters, since I was looking at the data using a software without a scale, it may have been more! Also, the results from multiple runs were very different, so the phone is not precise.
But these features you speak of, are they hardware or software features? (Sorry, I'm quite a n00b at these things!). Looks like GAGAN offers some good quality, but it's coming out in 2014!
I'm not looking at buying any unit as of now, was hoping to borrow or rent someone's! :P

[Srini]

[Sreepathi Pai]

On Sun, Dec 30, 2012 at 4:48 PM, Srini <srinidhi...@gmail.com> wrote:

> But these features you speak of, are they hardware or software features?
(Sorry, I'm quite a n00b at these things!). Looks like GAGAN offers
some good quality, but it's coming out in 2014!

Hardware + Firmware, most likely.

--
Sreepathi Pai

[Sunil Raghavan]

Interesting, I thought this thread would be a features discussion, but just saw its about accuracy.

If the discussion is about accuracy, my understanding: the only difference can be that one device fails to pick up some satellites at times where the other does. The contents of the satellite signals are digital and are used in similar algorithms across devices, so if the signal reception is the same, the location computation should be too. I've seen sites talking about one device being better than another at reception under trees, etc.

[Harpreet singh Wadhwa]

The accuracy of the Smartphone depends on the device, as many smartphones now a days don't have an actual GPS, they use internet connection to get positioning, in this case we get inaccurate date.

But if your device has actual GPS (in built), then it will give same results as Garmin. Coz they both work with same satellites.

Also Note that you are in India, so both devices may end up showing some inconsistency.

Thanks

On Mon, Dec 31, 2012 at 6:58 AM, Sunil Raghavan <raghava...@gmail.com> wrote:

Interesting, I thought this thread would be a features discussion, but just saw its about accuracy.

If the discussion is about accuracy, my understanding: the only difference can be that one device fails to pick up some satellites at times where the other does. The contents of the satellite signals are digital and are used in similar algorithms across devices, so if the signal reception is the same, the location computation should be too. I've seen sites talking about one device being better than another at reception under trees, etc.

--
biking conversations on the world famous "Bangalore Bikers Club" :)

are you a part of the bicycle racing scene?
Visit www.bangalorebicyclechampionships.com for more details

--

http://wharpreet.blogspot.com

http://linuxgazettes.blogspot.com
http://in.linkedin.com/in/wharpreet

[Banibrata Dutta]

The accuracy has at least the following 4 factors :

- Sensitivity of the GPS module (the hardware), i.e. how well it works when satellite signal strength is low. Increasing sensitivity in smaller space often requires more expensive electronic parts

- Support of features like WAAS which are meant to improve accuracy, and Assisted-GPS where mobile network helps get faster fix, after loss of satellite visibility. These features also add to the overall cost

- Tracking, mapping software itself and logic that deal with disconnections, data-loss etc. Often software use clever algorithm to "guess" (extrapolate, use heuristics, cached-data etc). the missing bits, when things go wrong, s.a. losing the GPS fix (on sufficient no. of satellites), or internet connectivity etc.
- Proximity of the GPS module (while operating) to other sources of "radio noise" in it's close vicinity

Most often all of these factors play a role, in accuracy... but there could be other factors as well.

Having said that, I have this notion (a rather strong belief, I'd say) that companies for whom a certain technology is their bread-n-butter, their central offering, they usually make the best products, s.a. Garmin for GPS, Nikon/Canon for Cameras, Leica/Carl-Zeiss for lenz etc. On the other hand, toys from a burger store, and chinese food in a darshini, usually don't do so well. On that note, a smartphone is often trying to cram too many features in a very small space, and thus make many compromises.

Happy new year to all.

regards,
Banibrata
http://www.linkedin.com/in/bdutta
http://twitter.com/edgeliving

[Karthick Gururaj]

Hi Banibrata,

I think you made a very good summary of the factors that can affect the accuracy/precision. But I don't necessarily agree with the later reasoning - essentially because economies of scale might mean that a high volume product (Smart phone) might be able to offer per quality per rupee than Garmin. Just may..

The question can be answered by an experiment in any case.. There are primarily two kinds of errors in a sensor like the GPS module (http://en.wikipedia.org/wiki/Observational_error):

a. Statistical bias - It is like a constant value added to each measurement

b. Random errors - when multiple measurements are being taken, how close are they to each other.

For example, if there is a bug in mapping software, it will introduce a bias for every measurement made; The measurements will all agree with each other (assuming the sensor is 100% "accurate/precise"), but all of them will have the same constant error. On the other hand, even if there are no bugs in the mapping software, the sensor can introduce random errors in each measurement - due to variations in signal strength of the received GPS signal (for e.g.).

One method to check for random errors (or how precise the instrument is) is to take repeated measurements over the same course - and see how the values vary.

In my case (using Nokia phone + Endomondo), I get a very consistent set of measurements from home-office route - irrespective of day-to-day variations of cloud cover. Note that I keep checking only the distance covered, not elevation gained/lost (ideally must check that as well). I keep the phone "outside", visible to the sky. If I keep it in my pocket, I see a marked difference.

As far as accuracy (statistical bias) is concerned, one hopes the manufacturer took care of that :) It is a matter of proper calibration.

Warm regards,
Karthick

[deepakvrao]

Elevation gained/lost is very inaccurate if measured by GPS.

[Sreepathi Pai]

On Tue, Jan 1, 2013 at 11:34 AM, Karthick Gururaj <karthick...@gmail.com> wrote:

Hi Banibrata,

I think you made a very good summary of the factors that can affect the accuracy/precision. But I don't necessarily agree with the later reasoning - essentially because economies of scale might mean that a high volume product (Smart phone) might be able to offer per quality per rupee than Garmin. Just may..

 

Let me flip this around somewhat. Garmin can amortize development of better GPS algorithms across its whole range of GPS devices (kind of like how Shimano et al. trickle down new technology from their higher-end gruppos to their lower end equivalents). Smart phone companies don't have that luxury and it's not their focus either.

Like Banibrata pointed out, a lot of the intelligence in tricky situations comes from the firmware/software. A dedicated GPS device can afford to make more intelligent tradeoffs between computational complexity of its algorithms, their power consumption, real-time behaviour and accuracy than a smart phone ever can. However I concede that most GPSes are good enough. It's only when you're trying to land a plane based on GPS info that things get tricky :)

As far as your testing of random errors in the app go, you might wish to observe the SD of the distance travelled as well, though it would be better to check for point errors (same points, same time).

--
Sreepathi Pai