Using control charts to monitor your car

This post may reveal a bit too much about the author, but I believe that the message makes it worth sharing.

In December of 2009, I purchased a 2010 Ford Escape 4wd, thinking that I might be moving out of Texas to the snowy northern US.  I did not end up moving, but I kept the vehicle.  As I have always done, I began tracking the operating expenses for my vehicle, which included tracking the fuel usage.  Many people consider this task to be a waste of time and a sign of my personal issues, but I have always believed in the benefit of data tracking at work and at home.

In May of 2012, I thought that I noticed a drop in my Gas Mileage, so that I decided to keep a close watch on it.  It was summer here in Austin, which means that the air-conditioning is running often and can impact the gas mileage.  When it did not seem to change as it cooled a bit, I plotted the data in an individuals chart and noted an apparent process change.  I realized that the change was a special-cause, not a common-cause issue with the temperature.

  • It is hot every summer, and it was hotter the prior summer than in 2012.  I saw no drop in prior summers.

My first thought was dirty fuel injectors, so that I added a fuel injector cleaner.  On my October 7th fill up, I used Seafoam brand, which seems to be the only one that has always helped, and my MPG appeared to return to past levels.

The chart below shows the MPG, including data after my corrective action!

Individuals chart of Miles per Gallon accounting for a special cause.

Miles per Gallon Individuals Chart

Individuals chart of Miles per Gallon accounting for a special cause

After this, I checked my buying history and found that the May date is when a new gasoline station opened that was not a common name brandbut had great pricing.  I started buying nearly all of my fuel there in May.  That brand of fuel may not have been the problem, but it correlated enough that I will not choose to save the few cents a gallon any more.

A two sample t-test provided a p-value of .oo1, which shows it was a significant change, even though it was not out-of-control on the individuals chart.

By the way, I think it is just luck that the before-and after-control limits are so similar.

Since I have shared this much, I will also share my tracking of fuel prices over the same period of time.  It has an odd pattern; trends seem to track to 1 Jan and 1 July.  I do not know why.

Price of gas since 2009

Price of gasoline in the US since 12/15/2009

During this time I have driven and purchased fuel from Texas to California with two trips from Texas to Ontario Canada. Two entries were in Canada.

Can you explain the changes in fuel prices in July and January?

My fuel-tracking program is an application on my iPhone called “Gas Log.” Additionally, I use an application called “Gas Guru” to find the best fuel prices in my area.

Enhanced by Zemanta
5 Comments

    Chris Wall

    May I also suggest a multi-vari to determine the influence of brands and travel types (local v highway)?

      Rick Haynes

      The Multivari chart is a great graphical tool to show differences when you want to include more than one x in the chart. Nearly every other visualization technique uses a single x and single y. Thanks for the suggestion.

    jesse hamilton

    Nice article for practical use. I sent it to others in Daltile and received replys.
    One said that the new station may have been selling gas with more ethanol and that may be somehting else for you to monitor (ethanol %).

    Mike Clayton

    Yes, ethanol added in WINTER usually..reduces milage to improve emissions.
    And Escape is a Flex Fuel vehicle, so no damage from ethanol…even up to 85% (E85) they say….and next year 15% will be pushed (E15 instead of E10). Many Texans hate the EPA of course. I simply think its dumb to make ethanol from corn using petroleum fertilizers, food and meat prices, and NOT solving the problem.

    You may find that cheap gas stations have OLD gas, so summer could have winter ethanol in it…but unlikely if turnover is good.

    Then there is DUST…as famous Texas issue. Fuel Filters?
    The “lower milage” was more stable…less variation.
    Was that due to type of driving last summer?
    No easy answers to any SPC story….but interesting.

    Ron Marchelletta

    Hi Rick – it was good to see you’re still ‘practicing’ what you preach. Just curious though. It appears the range was significantly reduced also after the May shift. That could be a good thing, right; remember, reduce the variation first! Anyway take care. Your friend.

Comments are closed.