What's more accurate? Speedo or GPS?

[craigeth]

When my speedo says I'm doing 83 or 84, my GPS says I'm doing 80.

So pretty much my GPS says my speedo is out by 4kms/h.

The GPS is about 2-3 years old. But I'm just wondering which is more accurate? I doubt the speedo on my 1 year old car would be faulty.

[Fitter]

My speedo is about the same. Got in dads XR6 turbo and surprise.. Speedo was 4km less then gps.

Looks like car speedos are generally out that little bit to make people think they doing XX speed but actually going 4-5kmh slower

[zedd82]

Car speedo's are set to be a couple of kms out. So if it says you are doing 80 you may really only be doing 78. It is setup this way on purpose so that if your speedo is sitting on the speed limit you won't be speeding.

It kinda covers the car manufactures from law suits in regards to this.

The ECU knows what speed you are really doing as it needs to for things like ABS, it just adjusts the readings on the speedo to be lower.

[Gleno]

my previous cars that I have owned, my VS commodore and 98 camry the speedo readings were on par with GPS. All other cars I have owned were 5-6% faster compared to the GPS. One wonders does this mean that when you clock up 100,000 kms, you have actually only done 94,000?? Something worth thinking about.

[craigeth]

This makes sense. Also makes sense why when I'm doing 80, everyone else seems to be doing 85.. Lol.

I think the odometer would still read accurate though. So 100,000kms would still be 100,000kms and not 94,000. It just takes you longer to get to 100,000kms cause youre traveling 4km/h slower than everyone else :p

[debonaire]

Yes I think odometer would be accurate - I know on the F3 in NSW for instance there are even kilometre markers where you test out your odometer to make sure it matches the signposts.

Of course all this is dependent on having your tyres pumped to the proper rolling diameter.

[Gleno]

yea true I'll be driving up the F3 on Friday so I'll test out the kilometre markers and let you know how accurate it is

[aussie027]

Below is the relevant ADR info. 1st bolded para summarizes the limits.

ADR-- Speedometers
Summary- Up until July 2006 the Australian Design Rules required new cars to have speedos that are accurate to within 10% of actual speed.
The current Rules disallow under-reading, and permit over-reading by up to 4kmh + 10%.

When vehicle manufacturers install a speedo at the factory, the speedo has been calibrated using data that assumes the vehicle is fitted with new standard tyres inflated to full pressure, on standard rims, and usually with the tyre circumference measured before the wheel is put on a vehicle or when the vehicle is on a hoist (no load on tyre). When they test the car for compliance with the ADR, the speedo will invariably produce an over-read result.
The current Australian Design Rules require speedos to be calibrated when the vehicle is unladen, fitted with normal tyres inflated to full pressure with an allowance for tyre heating:
"Unladen vehicle" means the vehicle in running order, complete with fuel, coolant, lubricant, tools and a spare wheel (if provided as standard equipment by the vehicle manufacturer), carrying a driver weighing 75 kg, but no driver's mate, optional accessories or load.
"Tyres normally fitted" means the type or types of tyre provided by the manufacturer on the vehicle type in question; snow tyres shall not be regarded as tyres normally fitted;
"Normal running pressure" means the cold inflation pressure specified by the vehicle manufacturer increased by 0.2 bar.

The current ADR prohibits any under-reading:
"5.3. The speed indicated shall not be less than the true speed of the vehicle. At the test speeds specified in paragraph 5.2.5. above, there shall be the following relationship between the speed displayed (V1 ) and the true speed (V2).
0 ≤ (V1 - V2) ≤ 0.1 V2 + 4 km/h."
This formula means that the vehicle's actual (true) speed must not be greater than the displayed speed. (Displayed speed minus true speed must be greater than or equal to 0, and less than or equal to 4kmh plus 10% of true speed). This means that if your vehicle's actual speed is 100kmh, the displayed speed is permitted to be anywhere between 100kmh and 114kmh.
Prior to July 2006 the ADR allowed ±10%. This means that cars sold new prior to 1 July 2006 could comply with the ADR even if the speedo under-read by 10%. Despite this being theoretically possible, due to the testing procedures and the reasons stated below it is unlikely to occur in practice.
The reason the ADR are relevent to the debate is that some people argue that if the ADR allows cars to have speedos that under-read by up to 10%, then surely our road laws must also allow 10% tolerance. Therefore, the argument goes, we should not be fined if we travel at 110kmh in a 100kmh zone because we are travelling within the tolerances allowed by the ADR. Such a situation would be quiet generous to motorists and would make it impossible for anyone to allege that they were mislead by their inaccurate speedo.
One problem with this argument is that it ignores reality. It assumes that speedos are under-reading by 10% when the truth is the vast majority of speedos over-read by about 3%. So in effect, it is allowing the majority of motorists the opportunity to drive at 10% above the speed limit. If the speed limits are there for a reason, then they would all need to be reduced by 10% to counter the effect of motorists increasing their speed above it.
If a motorist has a speedo that under-reads by any amount, then that person could try to argue that they have a defence of "honest and reasonable belief" in a state of facts which if true would make their conduct innocent. (See Proudman v. Dayman). Even if it were available, a Proudman v. Dayman defence is not dependant on the existence of the Australian Design Rule specification. A court is concerned with what your speedo actually displays, not what the ADR says it could display. If your speedo under-reads and there is no ADR speedo specification in existence, your defence would be exactly the same. The amount of tolerance specified in the ADR is irrelevant to any defence at law. Any driver who relies on the old ADR for their belief that their speedo was accurate would probably fail because the old ADR does not require a speedo to be accurate. It allowed it to be out by up to 10% so you should know it is out by 10%. And if you want to argue that the ADR is a "law" which applies to the State of Victoria, then you also need to argue that ignorance of the law is no excuse, and therefore you are deemed to have known that your speedo was up to 10% under-reading and you had no basis for assuming it was correctly displaying your speed at the time of the alleged offence.
If having an incorrect speedo were a defence to speeding then all you need to do is ensure your speedo is wrong and you will never have to pay a fine. If the inaccuracy of your speedo is merely a theoretical possibility, then that just turns a weak argument into a hopeless one because you are not addressing what in fact happened to you on the date of the offence.
For most of the 20th century there was no requirement that cars be fitted with speedos, seat belts and many other items we all take for granted these days.

[blade888]

Yes it is a fact that our speedos are off by 4~5km/h and it is confirmed that the odometers are still 100% accurate.

You can demonstrate this by traveling at 80km/h per speedo (use cruise control if you want) set the dash screen to view fuel consumption and average speed, then press and hold the info button on dash until it resets while staying at 80k/h on speedo. It will then give you an immediate speed reading of current speed traveling as per true sensor reading for the next few seconds before next average speed calculation tick.

The speed it'll read out will line up to your GPS and/or be 4~5km/h lower than the speedo.

I've done and demonstrated this a few times myself.

[Momengah]

From my lectures in Uni i remember them saying that GPS are controlled by atomic clocks. They're extremely accurate... something like accurate to 10 nanoseconds. While they can have a slight error in locating your exact position on the earth (due to its carrier wave nature and refraction etc) the error should be constant; speed calculations should therefore not be affected by it.

This means that when the gps calculates the speed based on the time and distance travelled it should have a pretty precise reading.

[JaCe]

That assumes you have GPS reception and your unit is correctly calibrated.

I think using an OBDII reader is a good way to get the ECU's reading of speed which should be most accurate. Failing that, the GPS option works better than the speedo- I used to use my MMCS to see this vehicle data.

[Jarryd]

Our speedos are out by 5% according to my experience.

So at 60 indicated speed you're actually doing 3km/h slower

70 = 3.5km/h slower = 66.5 true speed

80 = 4km/h slower = 76 true speed

90 =4.5km/h slower = 85.5 true speed

100 = 5km/h slower = 95 true speed

110 = 5.5km/h = 104.5 true speed

These speeds have all been constant with my GPS (Android Galaxy S using CoPilot) on straight, level roads (including highways and suburban roads).

I still tend to follow the speedo unless I'm in a hurry.

[RKC]

pretty much all manufaturers put the speedo a couple of ks faster than the actual speed. makes it so you cant blame them for the speedo being out if you get booked.

[billy boy]

assuming that you have the stock tyre/rim set up - once you change anything like that it will read differenlty, although you can get it recalibrated iirc. GPS is the way to go for the accurate reading