Car Fuel Efficiency Predictor Application Pitch

Course Project Presenter
2026-11-03

Slide 1: The Problem

With fluctuating fuel costs, consumers and manufacturers need simple, rapid tools to estimate a vehicle's fuel economy during early design or purchase phases.

  • Complex simulation software requires too many niche parameters.
  • Raw data tables are hard for non-technical users to digest.
  • Existing online tools often hide their baseline mathematical formulas.

Slide 2: The Shiny Solution

The Car Fuel Efficiency Predictor provides an intuitive, web-accessible tool to solve this problem dynamically.

  • Immediate Feedback: Watch predicted MPG shift instantly based on inputs.
  • Embedded Documentation: Novice-friendly instructions are built straight into the sidebar UI.
  • Visual Mapping: Outputs a real-time reactive graph mapping the user's specific vehicle against industry data standards.

Slide 3: Underlying Data & Model

The app uses a reliable multivariable linear regression model derived from the trusted R mtcars dataset, factoring in vehicle weight (wt) and transmission type (am).

fit <- lm(mpg ~ wt + am, data = mtcars)
summary(fit)$coefficients

               Estimate Std. Error     t value     Pr(>|t|)
(Intercept) 37.32155131  3.0546385 12.21799285 5.843477e-13
wt          -5.35281145  0.7882438 -6.79080719 1.867415e-07
am          -0.02361522  1.5456453 -0.01527855 9.879146e-01

When a user adjusts inputs in the application, the server.R backend executes a predictive matrix multiplication behind the scenes.

For instance, assuming a vehicle weight of 3,200 lbs (wt = 3.2) and an Automatic Transmission (am = 0):

example_input <- data.frame(wt = 3.2, am = 0)
predict(lm(mpg ~ wt + am, data = mtcars), newdata = example_input)

       1 
20.19255

Slide 5: Try It Today

Experience how straightforward data product interactions can be.

  • Open Source Codebase: Review the complete, reproducible repository containing both ui.R and server.R files on GitHub.
  • Local Replication: Users can run this application locally in RStudio using the open-source files provided in the repository.