- 87,462
- Rule 12
- GTP_Famine
It's worth reading beyond the headline.EA288 diesel engines don't have the cheat software:
http://www.honestjohn.co.uk/news/vo...ns-scandal-ea288-diesel-engines-not-affected/
It still won't do much for VW's reputation.
What VW actually said was "no software constituting an improper defeat device as defined in law is installed in vehicles with EA288". That's a very carefully worded statement which has passed over the corporate lawyers a few times in the process.
It's because it's not about what the cars do when actually driving... It's all about that New European Driving Cycle - and because this is a standardised test it has standardised procedures. All of the exhausts emissions are collected and analysed by weight and you can't do that on the road because you'd need to carry a bag behind you capable of holding about a million litres of gas.I really fail to see why it is so hard to stick a probe in an exhaust, plug in a laptop to the obd and go for a drive.
Exhaust probes, such as the ones you get on roadworthiness tests like the UK's MOT, measure small volumes of air for parts-per-million. That's largely good enough to assess how your car is performing compared to some relatively broad limits, but not sufficient for the finer-tuned tests that determine a car's tax bracket. When you consider VW's figure of £100m per 1g reduction in CO2, it's no surprise that manufacturers would not be keen on the "plug it in and drive it about" technique.
NEDC is at fault in many respects. It's a bad test that doesn't reflect the real world and the procedures are known and can be gamed. Few manufacturers go as far as to actually cheat it - I mean, who'd cheat on a test that you already know what the questions are?
A test needs to be developed that gets to within the 75th percentile of normal car use and the driving habits of the 75th percentile of drivers - with the test repeated for "doddering" and "lunatic" drivers to give a confidence level. As things stand, the first part of the test is "accelerate to 15km/h in 4s" and I don't even drive that lethargically in a traffic jam. This is also the most strenuously accelerative part of the test... And at no point is steering required.
Then the exact testing procedure needs to be concealed from manufacturers. They will not appreciate this, especially if their cars are found to be worse than someone else's in a test conducted in secret and so therefore obviously biased against them, so failing that it should be conducted at an automotive testing facility*, outdoors so that all four wheels are turning, at as close to 75th percentile weather as is possible (temperature, pressure, humidity). The test should include manouevring, so that some turn angle is involved to limit the usefulness of that to 'defeat devices'. It should also be repeatable to within 10% by a random selection of production models plucked from dealerships before sale.
* Of course if the GPS detects the location as a test facility, it might enable a 'defeat device' mode - but these facilities are not just for emissions purposes, so that would not be wise.