Wolfram Blog
Andrew Steinacher

Create a Tracker to Analyze Gas Mileage Using Wolfram Tech

June 22, 2017 — Andrew Steinacher, Wolfram|Alpha Developer, Wolfram|Alpha Scientific Content

Completed reportPlot 3D animation

When I first started driving in high school, I had to pay for my own gas. Since I was also saving for college, I had to be careful about my spending, so I started manually tracking how much I was paying for gas in a spreadsheet and calculating how much gas I was using. Whenever I filled my tank, I kept the receipts and wrote down how many miles I’d traveled and how many gallons I’d used. Every few weeks, I would manually enter all of this information into the spreadsheet and plot out the costs and the amount of fuel I had used. This process helped me both visualize how much money I was spending on fuel and manage my budget.

Once I got to college, however, I got a more fuel-efficient car and my schedule got a lot busier, so I didn’t have the time to track my fuel consumption like this anymore. Now I work at Wolfram Research and I’m still really busy, but the cool thing is that I can use our company technology to more easily accomplish my automotive assessments.

After completing this easy project using the Wolfram Cloud’s web form and automated reporting capabilities, I don’t have to spend much time at all to keep track of my fuel usage and other information.

Tracking MPG with Web Forms

To start this project, I needed a way to store the data. I’ve found that the Wolfram Data Drop is a convenient way to store and access data for many of my projects.

I created a databin to store the data with just one line of Wolfram Language code:

bin = CreateDatabin["Name" -> "MPG tracking"]

Basic Web Form

Next, I needed to design a web form that I could use to log the data to the Databin. I used FormFunction to set up a basic one to record gallons of fuel used (from filling the tank each time) and trip distance (from reading the car’s onboard computer).

I also added another field for the date and time of the trip, so that I could add data retroactively (e.g. entering data from old receipts).

I used the DateString function to create an approximate time stamp for submitting data:

basicForm = FormFunction[ { {"TripDistance", "Trip distance (mi)"} -> Restricted["Quantity", "Miles"], {"FuelUsed", "Fuel used (gal)"} -> Restricted["Quantity", "Gallons"], {"Timestamp", "Date & Time"} -> "ComputedDateTime", "Input" :> DateString[{"Month", "/", "Day", "/", "Year", " ", "Hour", ":", "Minute"}] |> }, (DatabinAdd[bin, #]; "Data submitted sucessfully!") & ]

This form works in the notebook interface, but it isn’t accessible from anywhere but my Mathematica notebook. If you want it to access it on the web or from a phone, you need to deploy it to the cloud.

Conveniently, you can do this with just one more line of code using CloudDeploy:

coBasic = CloudDeploy[basicForm, "CarHacks/MPGTracker/BasicForm", Permissions -> "Public"]

Wolfram Cloud form

Extended Form: More Data, Better Appearance

If that’s all you wanted to record, you could stop there. After just a few lines of code, the form created will log distance traveled and fuel used, but there’s quite a bit more data that is available while at a gas station.

A typical car’s dashboard shows average speed and odometer readings from the onboard computer. Additionally, most newer cars will report an estimation of the average gas mileage on a per-trip basis, so I designed the following form that makes it easy to test the accuracy of those readings.

I also added a field to record the location by logging the city where I am filling up with the help of Interpreter. I used $GeoLocationCity and CityData to pre-populate this field so I don’t have to type it out each time.

Finally, if you’re saving for college like I was, you’ll want to record the total price too.

All of these data points can be helpful for tracking fuel consumption, efficiency and more.

extendedFormSpec = {    {"EstimatedMPG", "Estimated gas mileage (mi/gal)"} ->      Restricted["Quantity", ("Miles")/("Gallons")],    {"AverageSpeed", "Average speed (mph)"} -> <|      "Interpreter" -> Restricted["Quantity", ("Miles")/("Hours")],      "Required" -> False,      "Help" -> "(optional)"      |>,    {"Odometer", "Odometer (mi)"} -> Restricted["Quantity", "Miles"],    {"TripDistance", "Trip distance (mi)"} ->      Restricted["Quantity", "Miles"],    {"TotalPrice", "Total price ($)"} -> "CurrencyAmount",    {"FuelUsed", "Fuel used (gal)"} ->      Restricted["Quantity", "Gallons"],    "City" -> <|      "Interpreter" -> "City",      "Input" :> CityData[$GeoLocationCity, "FullName"],      "Default" :> $GeoLocationCity,      "Required" -> False,      "Help" -> "(optional)"      |>,    {"Timestamp", "Date & Time"} -> <|      "Interpreter" -> "ComputedDateTime",      "Input" :>        DateString[{"Month", "/", "Day", "/", "Year", " ", "Hour", ":",          "Minute"}]      |>    };

The last thing to consider before deploying the webpage is the appearance. I set up some visual improvements with the help of AppearanceRules, PageTheme, and FormFunction’s "HTMLThemed" result style:

extendedForm = FormFunction[    extendedFormSpec,    (DatabinAdd[bin, #]; "Data submitted sucessfully!") &,    "HTMLThemed",    AppearanceRules -> <|      "Title" -> "My Car's Gas Mileage",      "ItemLayout" -> "Inline"      |>,    PageTheme -> "Red"    ]; coExtended =   CloudDeploy[extendedForm, "CarHacks/MPGTracker/ExtendedForm",    Permissions -> "Public"]

Advanced Wolfram Cloud form

Making It Accessible

Now that I have a working form, I need to be able to access it when I’m at a gas station.

I almost always have my smartphone on me, so I can use URLShorten to make a simpler web address that I can type quickly:

URLShorten[coExtended]

Or I can avoid typing out a URL altogether by making a QR code with BarcodeImage, which I can read with my phone’s camera application:

BarcodeImage[coExtended[[1]], "QR"]

Once I accessed the form on my phone, I added it as a button on my home screen, which makes returning to the form when I’m at a gas station very easy:

Add form to home screen

Visualizing and Analyzing MPG Data

If you’re following along, at this point you can just start logging data by using the form; I personally have been logging this data for my car for over a year now. But what can I do with all of this data?

With the help of more than 5,000 built-in functions, including a wealth of visualization functions, the possibilities are almost limitless.

I started by querying for the data in my car’s databin with Dataset:

binData = Dataset[Databin["me8Q5puO"]];

binData[[-5 ;;]]

With a few lines of code and the built-in entity framework, I can see all of the counties where I’ve traveled over the last year or so using GeoHistogram:

Show[  GeoGraphics[{EdgeForm[Black],     Polygon /@ {Entity[       "AdministrativeDivision", {"Illinois", "UnitedStates"}],       Entity["AdministrativeDivision", {"Indiana", "UnitedStates"}],       Entity["AdministrativeDivision", {"Wisconsin",         "UnitedStates"}]}}],  GeoHistogram[binData[[All, "City"]], "AdministrativeDivision2"]  ]

I can also see the gas mileage over the course of the past year with TimeSeries:

timeSeries = TimeSeries[Databin["me8Q5puO"]

DateListPlot[timeSeries["TripDistance"]/timeSeries["FuelUsed"],   PlotTheme -> "Detailed", FrameLabel -> {None, "mi/gal"},   PlotLabel -> "Gas Mileage", PlotRange -> Full, Mesh -> Full]

Analyzing Gas Mileage Factors

I often wonder what I can do to improve my gas mileage. I know that there are many factors at play here: driving habits, highway/city driving, the weather—just to name a few. With the Wolfram Language, I can see the effects of some of these on my car’s gas mileage.

I can start by looking at my average speed to compare the effects of highway and city driving and compute the correlation:

mpgVsSpeed =    Select[binData[[All, {"AverageSpeed", "EstimatedMPG"}]],     FreeQ[_Missing]]; ListPlot[mpgVsSpeed, PlotTheme -> "Detailed",   FrameLabel -> {Quantity[None, "Miles"/"Hours"],     Quantity[None, "Miles"/"Gallons"]},   PlotLabel -> "MPG vs Average speed"]

Correlation @@ Transpose[mpgVsSpeed]

It’s pretty clear from the plot that at higher average speeds, gas mileage is higher, but it does appear to eventually level off and somewhat decrease. This makes sense because although a higher average speed indicates less city driving (less stop-and-go traffic), it does require burning more fuel to maintain a higher speed. For example, on the interstate, the engine might be running above its optimal RPM, there will be more wind resistance, etc.

With the help of WeatherData, I can also see if there is a correlation with gas mileage and temperature. I can compute the mean temperature for each trip by taking the mean temperatures of each day between the times that I filled up:

binDataWithTemperature =    Dataset@BlockMap[(Append[Last[#],         "Temperature" ->          Mean[WeatherData[Last[#]["City"], "MeanTemperature",            Append[#[[All, "Timestamp"]], "Day"],            "NonMetricValue"]]]) &, Normal[binData], 2, 1];

mpgVsTemp =    binDataWithTemperature[[All, {"Temperature", "EstimatedMPG"}]]; ListPlot[mpgVsTemp, PlotTheme -> "Detailed",   FrameLabel -> {Quantity[None, "DegreesFahrenheit"],     Quantity[None, "Miles"/"Gallons"]},   PlotLabel -> "MPG vs Mean Temperature"]

The correlation is weaker, but there is a relationship:

Correlation @@ Transpose[mpgVsTemp]

I can also visualize both correlations for the average speed and temperature in 3D space by using miles per gallon as the “height”:

ListPointPlot3D[  Values@Select[    binDataWithTemperature[[     All, {"Temperature", "AverageSpeed", "EstimatedMPG"}]],     FreeQ[_Missing]],  PlotStyle -> PointSize[Large], AxesLabel -> {"Temp", "Speed", "MPG"},   Filling -> Axis, PlotTheme -> "Detailed",   PlotRange -> {All, All, {15, 40}}]

Plot 3D animation 2

It’s also clear from this plot that gas mileage is positively correlated with both temperature and average speed.

Automated Reporting

Now that I have code to visualize and analyze the data, I need some way to automate this process when I’m away from my computer. For example, I can set up a template notebook that can generate reports in the cloud.

To do this, you can use CreateNotebook["Template"] or File > New > Template Notebook
(File > New > Template in the cloud).

After following John Fultz’s steps in his presentation to mimic the TimeSeries plot above, I created a simple report template here:

Report template

I can test the report generation locally by using GenerateDocument (or with the Generate button in the template notebook):

SetDirectory[NotebookDirectory[]]; GenerateDocument["CarHacks1_BasicTemplate.nb", <|   "BinID" -> "me8Q5puO"|>, "CarHacks1_BasicReport.nb"]

Gas mileage report

From here, I can generate a report every time I submit the form by adding this code to the form’s action. But first I need to upload the template notebook to the cloud with CopyFile (alternatively, you can upload it via the web interface):

SetDirectory[NotebookDirectory[]]; CopyFile["CarHacks1_BasicTemplate.nb",   CloudObject["CarHacks/MPGTracker/BasicTemplate"]] SetOptions[%, Permissions -> "Public"];

Now I can update the form to generate the report, and then use HTTPRedirect to open the report as soon as it is finished:

automatedBasicReportForm = FormFunction[    extendedFormSpec,    With[      {       binID = "me8Q5puO",       templatePath = "CarHacks/MPGTracker/BasicTemplate",       reportOutputPath = "CarHacks/MPGTracker/BasicReport_Latest.nb"       },      DatabinAdd[Databin[binID], #];      GenerateDocument[templatePath, <|"BinID" -> binID|>,        reportOutputPath];      HTTPRedirect[CloudObject[reportOutputPath]]      ] &,    "HTMLThemed",    AppearanceRules -> <|      "Title" -> "My Car's Gas Mileage",      "ItemLayout" -> "Inline"      |>,    PageTheme -> "Red"    ]; CloudDeploy[automatedBasicReportForm, \ "CarHacks/MPGTracker/AutomatedBasicReportForm",   Permissions -> "Public"]

That is a basic report. Of course, it’s easy to add more to the template, which I’ve done here, incorporating some of the plots I created before, as well as a few more. Again, I can generate the advanced report to test the template:

SetDirectory[NotebookDirectory[]]; GenerateDocument["CarHacks1_AdvancedTemplate.nb", <|   "BinID" -> "me8Q5puO"|>, "CarHacks1_AdvancedReport.nb"]

Travel report

Seeing that it works, I can upload the template to the cloud:

automatedAdvancedReportForm = FormFunction[    extendedFormSpec,    With[      {       binID = "me8Q5puO",       templatePath = "CarHacks/MPGTracker/AdvancedTemplate",       reportOutputPath = "CarHacks/MPGTracker/AdvancedReport_Latest.nb"       },      DatabinAdd[Databin[binID], #];      GenerateDocument[templatePath, <|"BinID" -> binID|>,        reportOutputPath];      HTTPRedirect[CloudObject[reportOutputPath]]      ] &,    "HTMLThemed",    AppearanceRules -> <|      "Title" -> "My Car's Gas Mileage",      "ItemLayout" -> "Inline"      |>,    PageTheme -> "Red"    ]; CloudDeploy[automatedAdvancedReportForm, \ "CarHacks/MPGTracker/AutomatedReportForm", Permissions -> "Public"]

Lastly, I need to update the form to use the new template and then deploy it:

automatedAdvancedReportForm = FormFunction[    extendedFormSpec,    With[      {       binID = "me8Q5puO",       templatePath = "CarHacks/MPGTracker/AdvancedTemplate",       reportOutputPath = "CarHacks/MPGTracker/AdvancedReport_Latest.nb"       },      DatabinAdd[Databin[binID], #];      GenerateDocument[templatePath, <|"BinID" -> binID|>,        reportOutputPath];      HTTPRedirect[CloudObject[reportOutputPath]]      ] &,    "HTMLThemed",    AppearanceRules -> <|      "Title" -> "My Car's Gas Mileage",      "ItemLayout" -> "Inline"      |>,    PageTheme -> "Red"    ]; CloudDeploy[automatedAdvancedReportForm, \ "CarHacks/MPGTracker/AutomatedReportForm", Permissions -> "Public"]

With this setup, I can always access the latest report at the URL the form redirects me to, so I find it handy to also keep it on my phone’s home screen next to the button for the form:

Generate report

Conclusion

Now you can see how simple it is to use the Wolfram Language to collect and analyze data from your vehicle. I started with a web form and a databin to collect and store information. Then, for convenience, I worked on accessing these through my smartphone. In order to analyze the data, I created visualizations with relevant variables. Finally, I automated the process so that my data collection will generate updated reports as I add new data. Altogether, this is a vast improvement over the manual spreadsheet method that I used when I was in high school.

Now that you see how quick and easy it is to set this up, give it a try yourself! Factor in other variables or try different visualizations, and maybe you can find other correlations. There’s a lot you can do with just a little Wolfram Language code!


Download this post as a Computable Document Format (CDF) file. New to CDF? Get your copy for free with this one-time download.

Leave a Comment

One Comment


Gustavo Delfino

Great article. It would be great to continue the automation by interfacing with a wireless OBD2 tool.

Posted by Gustavo Delfino    June 28, 2017 at 12:38 pm


Leave a comment

Loading...

Or continue as a guest (your comment will be held for moderation):