BobSchuman

Incorrect on testing at different speeds. It would make the whole thing to convoluted. The problem with the way they do it now is that everyone knows how the epa figures mileage and they engineer the vehicles to take the test. What if the EPA figured out what we call "real world" and tested based on that?

You are right about the increased mileage and performance. However, people are still driving about the same speeds, so don't confuse these; they are separate issues. In other words, if you use the performance, you will get terrible gas mileage. The idea of "v8 performance and v6 economy" is misleading. What it should say is "or".

Ron AKA

EPA lets you compare vehicles and for each vehicle end users can submit their actual fuel economy for comparison. If enough users submit then you can get a pretty good estimate of whether or not actual users are achieving EPA or not. I randomly picked 4 F150 models from 2013, and you can find them at the link below. However the problem is that F150 users obviously do not submit many actuals. Of the four models only two had actuals, and the actuals were from one user each. Both actuals were below Ford's EPA, but two users is not statistically significant. You would have to fish around and try and find models that had more actuals reported to get anything useful.


http://www.fueleconomy.gov/feg/Find....33183&id=33186

Ron AKA

In other words you think it is fair to the vehicle purchaser for the manufacturer to design the vehicle so it gets better mileage on the EPA test, than the end user is ever going to achieve? I suspect the reason Hyundai/Kia got fined is that they outright cheated. Ford outright gamed the test, but did not technically cheat (at least in the case of the hybrids). The end result in both cases is the same, the end purchaser is promised something they cannot achieve. This may have saved Ford from an outright fine, but they know they deceived the buyer intentionally. In any case the cat is out of the bag, and Ford has been warned. That is the real reason for the EPA results delay for 2015. They know they are being watched this time.

BobSchuman

I think you are missing the point. The EPA test is very accurate. I consider the style of driving that they do on the test to be optimistic, but everyone drives the same course. The fact is that while the test is optimistic, there are people who get better mileage by driving slower and more conservative on acceleration and braking using gas-saving techniques.

The point of the EPA is to give people a basis for comparison. What we call "real world," is a more carefree and aggressive style of driving, which only serves to show that people don't really care about saving money on gas. What people really care about is how fast they can get where they are going.

I don't know the numbers, but the vast majority of people driving f150's could be driving a prius. I wish they would so that gas prices would go down with demand.

Jerry44

I don't think any manufacturer is designing a vehicle to somehow magically get better fuel economy on the epa test while getting relatively poor fuel economy in the real world. They design vehicles to get better fuel economy period. If it gets better fuel economy on the epa test, it is going to get better fuel economy in the real world, unless they gamed the test somehow. But if they gamed the test it wasn't through the design of the vehicle. They aren't going to design a vehicle around a test that could change at any moment. They are doing what they can to improve fuel economy. That's all there is to it. When you look at the design of the f-150, they obviously have done about as much as they can to improve fuel economy without making the vehicle too much more expensive. The 2.7 liter engine was designed to get the best possible fuel economy out of a gas engine while still having enough power for a truck. There isn't a lot left that they can do to improve fuel economy now without hybrid technology or alternative fuels.

BobSchuman

I don't think

Actually, all manufacturers including prius are looking closely at the test and engineer to get the best possible numbers. People have specifically complained about the practice of the methodology. I encourage you to look into this.

Jerry44

You have any source for that? I want to know how you can design a vehicle that is going to do relatively better on the test than in the real world. On every example I have heard of vehicles doing much worse in the real world, it was because they cheated with the actual test, not because they somehow cheated with the design.

Curmudgeon

It is documented in recent Gov't testimony re: Hyundai/Kia litigation that the manufacturer's mileage data submitted to EPA was tainted by testing vehicles in a tailwind, cherry-picking the most favorable numbers instead of averages, wrong tire sizes, and testing during warm weather which favored higher mileage numbers.

Ron AKA

How Ford did it with the Fusion was to use an artificial intelligence computer program. Artificial intelligence works by analyzing a whole bunch of data, and "learning" from it. A common example used is to feed a computer data on the sales of ice cream cones per day from an ice cream shop. You also feed it data on how hot it was, what day of the week it was, whether or not it was a holiday day, time of year, etc. With enough runs of data the computer artificial intelligence code forms it's own formula to predict what ice cream sales per day will be based on the temperature, day of week, etc. These programs have been proven to be very accurate, and are used by some retailers to determine how many staff they need and when.


In a Hybrid you have the ability to store gasoline energy in the batteries. The batteries are not really that big, so are constantly being charged and discharged. The overall objective is to keep the internal combustion engine (ICE) running in the most efficient zone, or have it off. For example if you have a hill coming up, it is most effective for the ICE to work a little extra hard, but still within the high efficiency zone to generate energy to be stored in the battery. Ideally you arrive at the bottom of the hill with a fully charged battery. Then when you climb the hill the battery power assists the ICE with the extra power needed. Again ideally the ICE stays within the maximum efficiency zone all the way up the hill. And if the computer knows there is a downhill right after the hill, it will fully deplete the battery and charge it back up again on the downhill kind of for free.


In real driving the computer can learn a little bit, but really does not know what is coming up. Eventually I suspect they will be married with GPS and know more, but currently I don't think they are smart enough to do that yet. But, and this is a big BUT, if it is an EPA test, the AI computer will perfectly learn the routine. It will know exactly when more and less power is needed, and it can prepare and operate the car perfectly. In short that is why Ford can game the EPA test. It has a computer that has already been "trained" to perfectly run the test.


Conventional vehicles without a battery storage system and electric motors do not have the tools to do this nearly as well, but the computer could still be trained to run the EPA test ahead of time.


And don't think for a second that Ford (and others) are not that sophisticated to do this. They know exactly how. EPA should force them to wipe the memory of these computers before every test, and let them learn as the test progresses each time. That replicates real world driving much more accurately.

11screw50

I have yet to own a vehicle that was not capable or exceeding EPA estimates. Those same vehicles were also capable of falling far short of the EPA estimates depending on how I chose to drive. I guess the EPA test could be changed to reflect that most people drive like morons most of the time...of course, there would then be no chance of meeting the target fleet economy levels without severely reducing available power. Personally, I would rather have more power and the choice of how I want to drive.