View previous topic :: View next topic |
Author |
Message |
CAA_1130221413 Lifetime Member
Joined: Nov 30, 2011 Posts: 5
|
Posted: Thu Feb 09, 2012 1:40 pm Post subject: wrong instant speed estimator |
|
|
The algorithm of instant speed estimator implemented in CamerAlert seems to use an integrator on the known speeds during the last 5-10 seconds. I can understand if it were computing an average speed of the last 1-2 seconds, but using such a broad time interval for estimating the instant speed will lead to delayed over speed detection (if car accelerates) or it will continue to issue over speed warnings although the car speed was brought under the legal limit (if car decelerates). Furthermore, I suspect it will also have an impact on the computed average speed accuracy.
IMO "GPS Test" application does a better job in estimating the instant speed. |
|
Back to top |
|
|
MaFt Pocket GPS Staff
Joined: Aug 31, 2005 Posts: 15258 Location: Bradford, West Yorkshire
|
Posted: Thu Feb 09, 2012 3:51 pm Post subject: |
|
|
Hi,
What device are you using? We've had a handful of reports about strange speed readouts but it's not affecting all devices...
For example on my Motorola Atrix it is fine...
MaFt |
|
Back to top |
|
|
DennisN Tired Old Man
Joined: Feb 27, 2006 Posts: 14901 Location: Keynsham
|
Posted: Thu Feb 09, 2012 6:03 pm Post subject: |
|
|
Hmm, interesting. I have always understood that during acceleration and deceleration, the current speed will not be accurate - I often get an overspeed indication as I enter a speed limit, even though I know for certain I'm under the limit by the time I enter it. But I have to say my van won't do 0-60 (or 60-0) in the sort of time that might cause me concern. _________________ Dennis
If it tastes good - it's fattening.
Two of them are obesiting!! |
|
Back to top |
|
|
lbendlin Pocket GPS Staff
Joined: 02/11/2002 22:41:59 Posts: 11878 Location: Massachusetts, USA
|
Posted: Fri Feb 10, 2012 1:32 am Post subject: |
|
|
"Instant speed estimator" - good one, I like it.
We're not estimating anything, we simply take what the GPS chip gives us. Crappy GPS=crappy data. If anything we may be up to one cycle behind in displaying the speed (ie we display the speed that the GPS had half a second before) - maybe that's what you are observing.
In the grand scheme of things this doesn't seem to impact the intended functionality. _________________ Lutz
Report Map Errors here:
TomTom/TeleAtlas NAVTEQ |
|
Back to top |
|
|
M8TJT The Other Tired Old Man
Joined: Apr 04, 2006 Posts: 10118 Location: Bexhill, South Sussex, UK
|
Posted: Fri Feb 10, 2012 9:27 am Post subject: |
|
|
I know that the SiRF Sar lll chipset has an 'averaging' mode (can't remember the real term) that I tried in my Mio C710. That introduced quite a lag on speed and position |
|
Back to top |
|
|
CAA_1130221413 Lifetime Member
Joined: Nov 30, 2011 Posts: 5
|
Posted: Fri Feb 10, 2012 2:06 pm Post subject: |
|
|
MaFt wrote: |
What device are you using? We've had a handful of reports about strange speed readouts but it's not affecting all devices...
MaFt |
I have an HTC Desire with Android 2.2.2. The problem is not that it estimates a wrong speed, but the fact that the convergence to real speed is really slow. For instance, if I drive with 50km/h and I suddenly stop, it will take at least 5 secs for CamerAlert to display an instant speed of 0 km/h.
lbendlin wrote: | "Instant speed estimator" - good one, I like it.
We're not estimating anything, we simply take what the GPS chip gives us. Crappy GPS=crappy data. If anything we may be up to one cycle behind in displaying the speed (ie we display the speed that the GPS had half a second before) - maybe that's what you are observing.
|
Then please explain why "GPS Test" application is able to print a pretty close to reality instant speed while speed reported by CamerAlert looks like it is given by an integrator (in other words, looks more like an average speed than an instant one). Even when I'm waiting at red lights, it still reports a speed of 5-10 km/h, speed that slowly converges to 0. |
|
Back to top |
|
|
Duddy Lifetime Member
Joined: Dec 09, 2006 Posts: 219 Location: Manchester UK
|
Posted: Fri Feb 10, 2012 3:25 pm Post subject: |
|
|
A speed calculation, based on GPS positioning, will always give an average speed.
Say the position is calculated every 1 second, the GPS will give the linear distance between the 2 GPS readings and carry out a distance over time calculation i.e if your are driving around a curve the calculation will be based on a straight line between the 2 points and slightly over estimate the speed.
When looking at the speed, shown at a GPS point, it will be the average speed over the previous second (look at Google earth .kml files as a good example).
The above shows that what you are reading as your current speed is actually your average speed for the previous second.
The speed readings become even more "average" if your GPS reads locations at greater than 1 second and even more delayed if you use a BT GPS receiver as the BT transmission takes an appreciable time to reach the speed processor.
If you look closely at a live mapping system on a laptop in a vehicle using a BT GPS you will see your location, on the screen, lags slightly behind your actual position.
GPS is great but you need to know it's limitations _________________ HUAWEI P30 Pro (new edition) with Speedtrap Alert & alcatel1 for SatNav
CoPilot 10 with CamerAlert
RoadHawk in-car video
Reading glasses getting thicker as is my waist
Retired but want to go back to work for a rest. |
|
Back to top |
|
|
M8TJT The Other Tired Old Man
Joined: Apr 04, 2006 Posts: 10118 Location: Bexhill, South Sussex, UK
|
Posted: Fri Feb 10, 2012 3:39 pm Post subject: |
|
|
Speed from the GPS chipset is not normally averaged between two position readings. It's a direct reading from thge satellite data and is called velocity. |
|
Back to top |
|
|
Duddy Lifetime Member
Joined: Dec 09, 2006 Posts: 219 Location: Manchester UK
|
Posted: Fri Feb 10, 2012 3:56 pm Post subject: |
|
|
M8TJT wrote: | Speed from the GPS chipset is not normally averaged between two position readings. It's a direct reading from thge satellite data and is called velocity. |
As I understand it the only signals, transmitted by a GPS Satallite are identification, time signal and correction factor.
Any location, speed or direction etc is post processing in the receiving device.
I am aware that speed and direction are a velocity but I tend not to use the term as some people are unaware of the difference between speed & velocity.
A good descrition of the GPS system is to be found in Wikipedia under the heading Basic concept of GPS
Edited to add the last paragraph _________________ HUAWEI P30 Pro (new edition) with Speedtrap Alert & alcatel1 for SatNav
CoPilot 10 with CamerAlert
RoadHawk in-car video
Reading glasses getting thicker as is my waist
Retired but want to go back to work for a rest. |
|
Back to top |
|
|
M8TJT The Other Tired Old Man
Joined: Apr 04, 2006 Posts: 10118 Location: Bexhill, South Sussex, UK
|
Posted: Fri Feb 10, 2012 5:21 pm Post subject: |
|
|
OK already. The signals transmitted by the GPS satelite are far more complex that you indicate, but the 'averageing' between two position readings a second apart does not give the averaging effect that the OP indicates.
As I said before, my GPS device (SiRF Star ll) has an averaging (smoothing)mode, which averages the speed and position over several 1 second readings. This gives the effect that the OP indicates i.e. speed and positional lag considerably more that the 'instantaneous' lag of an unsmoothed (usual type) output. |
|
Back to top |
|
|
lbendlin Pocket GPS Staff
Joined: 02/11/2002 22:41:59 Posts: 11878 Location: Massachusetts, USA
|
Posted: Sun Feb 12, 2012 3:01 pm Post subject: |
|
|
CAA_1130221413 wrote: |
Then please explain why "GPS Test" application is able to print a pretty close to reality instant speed while speed reported by CamerAlert looks like it is given by an integrator (in other words, looks more like an average speed than an instant one). Even when I'm waiting at red lights, it still reports a speed of 5-10 km/h, speed that slowly converges to 0. |
Ah, that is different. What CA does is to wait for significant GPS updates (rather than polling the position once every second). As a result it can happen that if you come to a sudden stop you may see CA still report the previous speed, since you being stopped now does not result in significant GPS updates. Fortunately the GPS chips are so crappy that even when you are stopped eventually the jitter will result in a significant update where your position jumps more than x meters, but your speed doesn't.
This is mostly a cosmetic issue, but we chose this approach to save energy. Waiting for data is more economical than polling. Especially when stationary. _________________ Lutz
Report Map Errors here:
TomTom/TeleAtlas NAVTEQ |
|
Back to top |
|
|
253 Lifetime Member
Joined: Mar 05, 2007 Posts: 1058 Location: The green bit between the M40, M4 and M25.
|
Posted: Mon Feb 13, 2012 1:19 am Post subject: |
|
|
A little addition, and mainly because of what CAA_1130221413 posted because it may / may not be relevant.
The HTC Desire on 2.2.2 is very low on internal memory (I have one) and no longer use it for navigation / CA etc for that reason. I have a Sensation instead.
That may be part of the problem, I don't know, just saying.
Bob _________________ Triumph Tbird 1700. And now a Bonnie T100. |
|
Back to top |
|
|
CAA_1130221413 Lifetime Member
Joined: Nov 30, 2011 Posts: 5
|
Posted: Mon Feb 13, 2012 9:31 am Post subject: |
|
|
lbendlin wrote: | This is mostly a cosmetic issue, but we chose this approach to save energy. Waiting for data is more economical than polling. Especially when stationary. |
Wouldn't make more sense to prioritize accuracy over power consumption? I think most users connect their devices to the lighter socket while driving, isn't it?
Or at least make it configurable... |
|
Back to top |
|
|
Darren Frequent Visitor
Joined: 11/07/2002 14:36:40 Posts: 23848 Location: Hampshire, UK
|
Posted: Mon Feb 13, 2012 9:39 am Post subject: |
|
|
Two issues there, firstly, this issue doesn't affect the majority, and secondly, it doesn't affect the accuracy of the speed measurement when moving.
As Lutz has said, it's purely cosmetic and even as you have described it, this has no affect on the speed calculation when you are moving at more than a few miles or kilometres per hour. _________________ Darren Griffin |
|
Back to top |
|
|
M8TJT The Other Tired Old Man
Joined: Apr 04, 2006 Posts: 10118 Location: Bexhill, South Sussex, UK
|
Posted: Mon Feb 13, 2012 9:53 am Post subject: |
|
|
Then do like TomTom do when they have a minor problem (or not) with a feature. Remove the facility all together |
|
Back to top |
|
|
|
Posted: Today Post subject: Pocket GPS Advertising |
|
|
We see you’re using an ad-blocker. We’re fine with that and won’t stop you visiting the site.
Have you considered making a donation towards website running costs?. Or you could disable your ad-blocker for this site. We think you’ll find our adverts are not overbearing!
|
|
Back to top |
|
|
|