GM's OLM (Oil Life Monitor) System

[Matt]

After a recent purchase of a Chevrolet Avalanche, I learned about this Oil Life Monitor used by GM to track oil life.

What it does is check specific variables like temperature, moisture, etc and, using an algorithm, calculate the "life expectancy" of your oil. The result is a 0-100% read-out as you drive, depending on driving habits, speeds, distances, etc, that tells you when to change your oil.

What do you think of this? Do you think all cars should switch to this method? Do you think it's a BS method used to get you to change oil more/less often?

I've noticed lots of people comment that it actually makes them feel more secure about when to get their oil changed, as the "every 3000 mile" rule shouldn't apply to anyone.

Does anyone know if the algorithm is changed at all when switching to synthetic?

[WarmAndSCSI]

It's a bunch of BS. Any system that provides advice based upon anything but empirical data is a waste.

The only way to tell when you "should" be doing oil changes is by doing a UOA every single change.

[Matt]

Originally Posted by WarmAndSCSI

It's a bunch of BS. Any system that provides advice based upon anything but empirical data is a waste.

The only way to tell when you "should" be doing oil changes is by doing a UOA every single change.

From what I've read, many people believe that the engineers that designed this system is pretty much an ongoing, dynamic UOA that changes along with your driving habits.

Of course, many believe it's ******** as well, haha. My friend changes his oil every 3,000 miles whether the OLM reads 10% or 90%.

Me? I'm going to see how the OLM reads my own driving habits first. If it routinely has me changing the oil every 2-4000 miles, I may lose faith in it, especially if a UOA proves I could be going longer.

[ChrisSlicks]

The problem with the GM/Ford calculations is that it doesn't know what kind of oil how have in there. Dyno and synthetic result in the same OLM numbers.

The GM system will typically have you go 5000-7000 miles before the OLM states you need an oil change.

[FricFrac]

...as technology advances I won`t be surprised to see a mass spectrometer on a single IC for $5 a pop... it will eventually happen and stuff like this will be normal....

[AuburnZ]

I used it on our tahoe that we just traded in on the nismo, it had 73k miles when we traded it in. I routinely went 7-8k miles before changing the oil. I usually changed when it got down around 25% (sometimes earlier when we were about to take a trip and tow the 07 350z show car).

[Chriz]

Does the Z use somthing like this? I know that the car sends out warnings when an oil and filter change are required but im not sure if its by millage or actual ecu readings.

[Matt]

Originally Posted by Chriz

Does the Z use somthing like this? I know that the car sends out warnings when an oil and filter change are required but im not sure if its by millage or actual ecu readings.

The Z is just a mileage counter. You can put in how many miles to warn you after.

[Jordo!]

Originally Posted by Matt

What it does is check specific variables like temperature, moisture, etc and, using an algorithm, calculate the "life expectancy" of your oil.

It really all depends on how well the algorithim models emprical data... it won't be 100%. That said, if it does better than a human staring blankly at a dipstick, it might be useful.

However, if you can't even select something to indicate dino vs synthetic, that adds in a lot of potential error right there...

I'd like to see how they developed the tech, the theory underlying the model, and how well their predicted values map onto empirical UOA's for a given oil, but my first blush response is it's nice idea but an even better gimmick.

[Matt]

Originally Posted by Jordo!

It really all depends on how well the algorithim models emprical data... it won't be 100%. That said, if it does better than a human staring blankly at a dipstick, it might be useful.

However, if you can't even select something to indicate dino vs synthetic, that adds in a lot of potential error right there...

I'd like to see how they developed the tech, the theory underlying the model, and how well their predicted values map onto empirical UOA's for a given oil, but my first blush response is it's nice idea but an even better gimmick.

The OLM has a specific set of constants to use during its calculations. One of those constants is the type of recommended oil. In cars like the CTS-V, the OLM is actually designed to only presume synthetic is in the car. On a Cobalt, it's always gonna presume there's dino in there.

But you're right, when a cobalt owner switches to Mobil 1, a common solution is to go through 2 cycles of the OLM, which pretty much defeats the whole purpose of system. (and should even make you question its validity)

Here are some of the variables that systems such as the OLM use to determine the remaining "life" of your oil:

Conductivity -- how easily electric current passes through the oil (typically, the lower the electrical resistance, the more contaminants are in the oil)

Mechanical properties -- piezoelectric sensors can tell how thick the oil is by the force feedback it gives when sloshing around

Soot concentration -- dirty oil's days are definitely numbered

Presence of water -- water is an impurity in oil, since it hampers the oil's effectiveness and can corrode metal surfaces

[Jordo!]

Originally Posted by Matt

The OLM has a specific set of constants to use during its calculations. One of those constants is the type of recommended oil. In cars like the CTS-V, the OLM is actually designed to only presume synthetic is in the car. On a Cobalt, it's always gonna presume there's dino in there.

But you're right, when a cobalt owner switches to Mobil 1, a common solution is to go through 2 cycles of the OLM, which pretty much defeats the whole purpose of system. (and should even make you question its validity)

Here are some of the variables that systems such as the OLM use to determine the remaining "life" of your oil:

Huh -- neat. Well, if it can get meaningful data from those tests then, yeah, it might work pretty well after all, even if you switch types of oil.

I guess one thing you could do is take an oil sample, send it for a UOA, and then compare and contrast. If the car's analysis maps on well to the UPA, you're probably good to go from that point on.

WarmandSci, what do you think?

[FricFrac]

Conductivity -- how easily electric current passes through the oil (typically, the lower the electrical resistance, the more contaminants are in the oil)

Mechanical properties -- piezoelectric sensors can tell how thick the oil is by the force feedback it gives when sloshing around

Soot concentration -- dirty oil's days are definitely numbered

Presence of water -- water is an impurity in oil, since it hampers the oil's effectiveness and can corrode metal surfaces


==================================================



Then it's a great idea. I would still use the regular oil change interval but this will show you if there is some other factor that has pre-maturely degraded the life of your oil.

[Mike]

well, all I know is when I towed frequently with the tahoe, it was about 7000 miles between oil changes, and in the time periods where I did no towing, it was about 13000, so it does know how the car is being used. As for the oil, like the one guy said, its based on what the recommended oil is, so if you up or down grade, results may vary. I don't keep a car much past warranty anyway, so I do the minimum required (except on the Z, as it gets an oil change every other track day, or about 6 or 7 changes per year.)