What is correct, Speedo or sat nav?

benjus

C_Mababejive wrote: »

In my view,the speedo is more accurate since it is more closely connected to the medium that is to be measured

It's physically closer, but not directly related. It can only measure the speed that the wheels are rotating; how this translates to the car's forward speed depends on the size of the wheels and tyres, which can vary.

I've certainly never been flashed by a camera while maintaining my speed under the speed limit using the SatNav's speed readout.

missile

C_Mababejive wrote: »

....
Anyway the police are interested in your speedo,not your sat nav.

The police are interested in your speed not how you measure it.

thescouselander

missile wrote: »

Only 1 mph difference between my Audi Speedo and my TomTom sat nav at 70 mph.

Same here. My old BMW was only 1mph out too. I guess the german manufactures can be bothered to calibrate their speedos properly.

Kilty_2

Be aware that a sat nav is only more accurate if it has a number of satellite connections (ie a strong signal) - most units let you see this though.

molerat

My speedo is a linear 3-4mph faster than the sat nav depending on tyres. My digital speedo (look up ford digital speedo) is much closer to the sat nav. Different sat navs have given the same readings. On steep hills and winding roads the sat nav will be less accurate as you are covering less linear distance over the earth's surface.

ScoobySnacks79_2

The answer is that neither is likely to be entirely accurate.

The speedo varies with tyre wear, and whether the tyres are correctly inflated. GPS is only accurate to a few dozen metres, and so can only give an average speed over a number of readings. For various reasons it is inherently slightly inaccurate.

Assuming that the speedo reads in a linear manner, then doubling the speed should double the difference between the speedo and the gps.

Most sat navs only use four satellites, and those on a step basis, i.e. reading form one satellite, reading form the next, reading from the next etc. This means that it is comparing readings at slightly different times. An aircraft GPS has several receivers (up to 12) so compares the readings at the same time and is thus more accurate.

NeedsAdvice

if you had 2 identical cars, one with new tyres and one with old tyres, the car with new tyres will cover a slightly greater distance with each rotation of the wheel than a car with old tyres, therefore would be travelling slightly faster, but the speedos would read the same.

I find the speedo in my Merc to be very accurate in relation my satnav and find myself overtaking through speed cameras as everyone else slows down to what they think is the correct speed.

I would say Sat Navs are pretty accurate, on straight flat roads.

cyclonebri1

NeedsAdvice wrote: »

I find the speedo in my Merc to be very accurate in relation my satnav and find myself overtaking through speed cameras as everyone else slows down to what they think is the correct speed.

I would say Sat Navs are pretty accurate, on straight flat roads.

Yep, I agree, and trust in my SN. +4mph readout from my speedo and does make a real differrence when travelling mile after mile on 50mph roads

Strider590

The GPS should be better than 99% accurate, it's even been used in court against a speeding fine and succeeded.

The system was originally developed by the military for accurately guiding weapons to targets and during the first Iraq war, GPS globally was tweaked to provide positioning accuracy of less than 3 metres over ANY terrain (originally 6 metres).
Since then it's accuracy improves with every GPS satalite added to the network (24 or so now).

Speed calculations are an average taken over a number of measurements (per millisecond), any discrepancy is quickly removed and the average is as accurate as it can get.

It's accuracy cannot be questioned.

NeedsAdvice wrote: »

I find the speedo in my Merc to be very accurate in relation my satnav and find myself overtaking through speed cameras as everyone else slows down to what they think is the correct speed.

I get this all the time on dualcarriageways where people think the limit is 60mph, I end up either sailing past at 70 or flashing my lights/honking at the pillock in front for slamming on the brakes to go past the camera!! :mad:
One camera in particular is on a bend, on several occasions the idiot in front see's camera, slams on brakes, loses control momentarily and starts to drift toward the central reservation (and the camera), at which point your life is flashing before your eye's because you've no choice but to do the same thing (unless you can switch lanes).

When your behind someone who's doing 70, you don't really expect them to do this.... Obviously they know (think) they're speeding!!!!!!! 100% idiots!

epninety

Strider590 wrote: »

It's accuracy cannot be questioned.

The system was originally developed by the military for accurately guiding weapons to targets and during the first Iraq war, GPS globally was tweaked to provide positioning accuracy of less than 3 metres over ANY terrain (originally 6 metres).
Since then it's accuracy improves with every GPS satalite added to the network (24 or so now).

Speed calculations are an average taken over a number of measurements (per millisecond), any discrepancy is quickly removed and the average is as accurate as it can get.

It cannot be questioned (maybe) but it can be misunderstood. GPS accuracy is not (and cannot be) a hard limit, it's an uncertainty with a confidence interval, not an absolute accuracy. That is to say, you can not say that the position accuracy is x, only that (e.g.) 95% of positions will be within x.

Measurements are taken with very high precision (in the nanosecond range) but position fixes are usually calculated only once per second in most consumer grade units. Faster solutions are available, 10Hz is becoming more common, 100Hz is available in commercial units if you have upwards of 10k to spend.

A few back of the envelope sums, for interest only...

Most consumer type GPS seem to read out the speed as 70mph, not 70.4mph so the actual speed could vary by 1mph (from 69.51mph to 70.49mph) without the display changing.
Therefore, in this case the accuracy cannot be better than 1.4%, even with a perfect GPS engine behind it.

Instant GPS accuracy can be off for all sorts of reasons, not all of them obvious. e.g. the GPS calculation could be low by 1% in a high speed corner (because the speed calculation cuts the corner as a series of straight lines)

On the other hand, my speedo reading could change by 3.5% due to tyre wear alone (using BF Goodrich own specs for their tyres). Not all manufacturers tyres are the same size, especially in the larger 4x4 sizes. Differences of that much again due to the manufactured size are quite possible.

But the truth is, GPS is usually better than any other 'general' method for measuring speed. But that is why there are (or should be) tolerances on enforced speed limits - to remove the effect of measurement uncertainty. If you get a ticket for driving at 35mph in a 30 zone, the limits should ensure than when all uncertainties (yours and theirs) are summed, you were still speeding.

Oh, and while typing this, I have a high spec GPS unit on test, plotting measurement variation in a real world situation. Due to a building in the way, the GPS has an obstructed view of the sky to the south, which reduces accuracy due to reducing the available satellites, and also by effectively placing all the satellites in one portion of the sky. I have an E/W variation of sub 10m, but a N/S variation of 40m or more.