http://www.engadget.com/2007/11/12/sonys-nv-u73t-and-nv-u83t-gps-units-hit-the-fcc/
Engadget wrote:
We'd already seen Sony's newest nav-u GPS units pop up in Europe, but from the looks of the FCC's all-knowing database, the NV-U73T and 83T are coming Stateside soon. New specs in the filing are hard to come by, but everything we already knew is there -- baby-brother U73T sports a 4.3-inch display, while the U83T gets a 4.8-inch screen, Bluetooth, and an accelerometer to allow the unit to keep the guidance going even when the GPS signal fades. No word on pricing or availability, but you know we just love the pretty pictures anyways.
So, is that feature N95able?
why both if you have A-GPS ? (ok so it costs a bit but data is now quite cheap)
because AGPS only helps you get an initial lock, not constant signal
That is quite ridiculous reading that.
Think about it.. how does an accelerometer that measures acceleration / deceleration of an object (in particular directions) have any effect of GPS? What happens if your car is travelling on constant velocity? This means that there will be nothing detected..
I suspect this may be a software implementation of an accelerometer rather than an actual hardware device. I would assume it works out your velocity while it has a lock and then uses that to estimate your movement while the lock is lost and being re-established.
using the accelerometer for the GPS is an extremely good and common feature in the industry. my built-in acura navigation system has its own gyroscope for accurate turns (even without a strong gps signal) and also measures tire rotation speed. this is good for example when you go thru a long 3 mile tunnel and lose all gps signal completely. if nokia were to implement this into the n95 firmware it would greatly increase acuracy when making a turn, because the software would anticipate the turn based on readings rather than wait until AFTER the turn and calculate based on new location after said turn. most high end navi systems do this already, so if nokia were to add this it would be great. however, this hardware has been sitting in this phone 9 months with no development, so dont expect nokia to magically "get it" now if they didnt from the beginning.
bchliu wrote:That is quite ridiculous reading that.Think about it.. how does an accelerometer that measures acceleration / deceleration of an object (in particular directions) have any effect of GPS? What happens if your car is travelling on constant velocity? This means that there will be nothing detected..
That's the general idea - No change of direction/speed = you're travelling in the same direction at the same speed, what will be visible on the satnav map. If there are any changes the software will calculate it using data from the accelerometer and ROUGHLY will be able to show your location on the map.
RogerPodacter wrote:using the accelerometer for the GPS is an extremely good and common feature in the industry. my built-in acura navigation system has its own gyroscope for accurate turns (even without a strong gps signal) and also measures tire rotation speed. this is good for example when you go thru a long 3 mile tunnel and lose all gps signal completely. if nokia were to implement this into the n95 firmware it would greatly increase acuracy when making a turn, because the software would anticipate the turn based on readings rather than wait until AFTER the turn and calculate based on new location after said turn. most high end navi systems do this already, so if nokia were to add this it would be great. however, this hardware has been sitting in this phone 9 months with no development, so dont expect nokia to magically "get it" now if they didnt from the beginning.
But thats completely different from an accelerometer. As you say, yours is built in and thus calibrated to the car, not to mention have something to measure tire rotation speed.
Personally I dont even remotely see how the accelerometer in the N95 can help with GPS. Different cars would have different effects, for example if you go round a corner in an old micra, compared to a high range sports car. Perhaps an accelerometer coupled with other things, such as gyroscopes, but not on its own.
Plus Im betting it would be resource intensive.
bartmanekul wrote:
Plus Im betting it would be resource intensive.
Sure is!
Have you been trying out the beta Rotate Me 2? theres a substantial battery life cut when having that in memory as it constantly polls the accelerometer for changes. Even Rolling ball tends to chew up battery.. hahaha
bartmanekul wrote:But thats completely different from an accelerometer. As you say, yours is built in and thus calibrated to the car, not to mention have something to measure tire rotation speed. Personally I dont even remotely see how the accelerometer in the N95 can help with GPS. Different cars would have different effects, for example if you go round a corner in an old micra, compared to a high range sports car. Perhaps an accelerometer coupled with other things, such as gyroscopes, but not on its own.
Plus Im betting it would be resource intensive.
The correlation i was trying to make was that it could mimick the gyro in my current system. The type me car wouldn't matter. If the accelerometer detects the car turning it would assist the turn on the gps screen. Wouldn't matter the exact speed of a particular car turn. The gyro does essentially the same thing now. But this would kill battery quicker. But hey using your phone as gps isn't exactly battery efficient as is now is it?
A digital compass would be far better for that.
Along with an accelerometer, it also lets you potentially do things like MARA, which while a bit "unfinished" shows some interesting ideas.
Imagine the advertising that Google could place with an application like that!
RogerPodacter wrote: The correlation i was trying to make was that it could mimick the gyro in my current system.
I seriously doubt that would work. Using an accelerometer in a phone to mimick a gyro? Doubtful if you ask me.
RogerPodacter wrote:
The type me car wouldn't matter. If the accelerometer detects the car turning it would assist the turn on the gps screen. Wouldn't matter the exact speed of a particular car turn. The gyro does essentially the same thing now.
So there wouldnt be any problems with the sensitivity then? Doubtful again. A gyro is aquite a bit different than an accelerometer, especially one in a phone.
RogerPodacter wrote:
But this would kill battery quicker. But hey using your phone as gps isn't exactly battery efficient as is now is it?
Why not?
Satnav on an N95 can usually last about 2 hours from a full charge, and thats with it being used constantly. Mine lasts about the same, maybe longer (a 6110).
Most dedicated satnavs dont last longer than that, and very few people dont plug satnav into some sort of power source. Thats what you expect to do with satnav really.
RogerPodacter wrote:using the accelerometer for the GPS is an extremely good and common feature in the industry. my built-in acura navigation system has its own gyroscope for accurate turns (even without a strong gps signal) and also measures tire rotation speed. this is good for example when you go thru a long 3 mile tunnel and lose all gps signal completely. if nokia were to implement this into the n95 firmware ...
Don't underestimate the N95 software (and Nokia Maps): Today I drove through more than a dozen of tunnels, some of them more than 3km long. The gps signal was lost completely in all of them. But by my surprise the gps position was constantly updating. The new Nokia Maps software (I have paid for navigation), has this feature implemented! :icon14: :icon14: I suppose that it takes the last known speed and continues from there to calculate the movement, following the road you're in. Several of the tunnels had turns and the map view was rotating accordingly.
And generally in the turns I have no problem at all. It behaves correctly always in all roundabouts.
I don't really see the point anyway, tbh. If you go into a tunnel with only one way in and out, what does it matter if the SatNav stops until you get out, you aren't going to get lost! In fact, even if there are turn offs in the tunnel, without a satellite link, no amount of clever gyros etc will know which way you have gone until you get out anyway
ph7 wrote:I suppose that it takes the last known speed and continues from there to calculate the movement, following the road you're in. Several of the tunnels had turns and the map view was rotating accordingly.
Tomtom does exactly the same, and like you I think it just calculates the distance from the speed when you go into the tunnel.
You could test it by going into the tunnel slowly, wait until the signal is lost, then speed up and see if the sat nav then goes out of sync with the road. I don't go through tunnels very often though to try it.