How accurate is Sat Nav speed readings

david29dpo

Post 8 tells you why.

Hintza

Well I use the Sat Nav for travelling through average speed cameras. Slight innacuracy on hills but not worth writing home about.

Grade_A_Reject

I was always under the impression that car manufacturers deliberately set their speedo readings higher than the actual speed. The reason being if you got prosecuted for speeding and said that your speedo was showing 70mph, for instance, and tried to sue them for the fact that you were driving at what their car said was a legal speed they would have better legal recourse, as if it was indicating 70mph then your actual speed would have been around 66mph.

Pssst

david29dpo wrote: »

Post 8 tells you why.

Its not really scientific proof is it? My mates a copper..bah blah blah.

Lorian

Time how long it takes to do a mile at a constant 60 miles an hour according to the GPS and the speedo and compare the results. It should of course take exactly 60 seconds.

In my car the speedo and GPS both agree at 0 mph. By the time the speed has increased to 60mph the speedo is reading low by about 6%. The speedo would be more accurate with new front tyres.

ziggyman99

Pssst wrote: »

Why?.....................

Most GPS manufacturers quote an accuracy of 0.1% in a good signal area at a constant speed. The only error would be due the fact that the display doesnt update immediately. The reason GPS is accurate is because the US military made is so! Thet need to know that their cruise missiles are going to arrive at the place they sent them to - and as the pictures showed in Iraq they did with deadly accuracy. The satellites all contain an atomic clock (caesium) which maintains an accuracy of 10-9 seconds per day. That's a second every few million years. Your GPS unit knows where the satellites are and knows the difference it takes for the signal to get to you, then works out the speed and location from that. That's putting it a little simplistic but it's basically correct.
So, yes a GPS unit is VERY accurate. Far more so than any normal car speedo.

BillScarab

The accuracy of your speedo depends on the amount of wear on your tyres, the tyre pressure and whether you have the correct size tyres fitted.

Here's a quote from a lorry driver taken from Honest John's website who compared, Sat-Nav, Speedo and Tacho graph.

"Using my satnav in both lorry and car I find that while its speed figures just about match those of the lorry’s calibrated tachograph a car’s speedo is usually about 10 per cent slower: ie 70mph on the speedo reads 63mph on the satnav. Asking around other lorry drivers who use their satnavs in both lorry and car comes up with similar results."

The one time a Sat Nav may be less accurate is in very hilly areas but only by 1-2%.

If you do a google search there is lots of discussion and evidence about this.

ziggyman99

BillScarab wrote: »

The one time a Sat Nav may be less accurate is in very hilly areas but only by 1-2%.

Hence my bit about a good signal. The GPs system uses triangulation so needs to see at least 3 satellites to work properly.

epninety

Pssst wrote: »

Its not really scientific proof is it? My mates a copper..bah blah blah.

Try thinking about it for a second. Your GPS knows you are HERE, NOW. A few moments ago, it knew you where THERE, THEN. It knows the position fairly accurately, and the time very accurately. If both position fixes are valid, then it's gcse maths to work out the average speed between the two, to a high accuracy.

The innacuracies come (disregarding bad position fixes)mainly from the inherent error in the calculated positions. GPS doesn't know exactly your position, it knows fuzzy area where it is much more likely that you are in the middle than at the edge.

Possibly the bigger error is the +/- 1 digit in the speed display. e.g. Huskie said the GPS indicated 61, and the speed gun 60, which is nearly a 2% error. But if the real speed measured by the GPS was 60.5, and the speed gun measured 60.4, then the real error is only 0.17%, only the results presented to the user have a larger error. For any digital display the limit of displayed accuracy is +/- 1 in the smallest digit.

photome

I hope by now that Psssst is convinced.