Tymoteush / Truck-transport-industry

Truck transport US industry analysis and diesel price forecasting model using ML and traditional Data Science approach.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Truck transportation industry project - United States

In this project I have analyzed data about Truck Transportation industry and related factors in the US for period 2006-2019 and performed forecasting of Diesel prices in the Gulf Coast region(PADD 3). The data used in this project is publicly available on the following websites:

Bureau of Transportation Statistics https://www.bts.gov/

Federal Reserve Bank of St. Louis https://fred.stlouisfed.org/

U.S. Energy Information Administration https://www.eia.gov/

The analysis and forecasting model will be used by Arkansas-based company offering For-Hire truck transportation services. The model will help company managers make decisions about diesel storage and ordering strategies.

The Company

The company currently operates 68 Peterbilt 389 trucks with Cummins X15 engines. The company has its own gas pump and underground tanks for diesel storage. They are ordering diesel from an external supplier. The trucks make around 450 miles per day and they return to the company headquarters on a daily basis. They drive 6 days a week for 51 weeks a year. The trucks are refueled only at company headquarters. The trucks average fuel consumption is around 6.5 miles per gallon.

Project use case

The company would like to be aware of industry trends and have insights about the overall direction the industry is heading in order to be able to compare overall industry performance to their own performance and evaluate business development strategy and goals. Average 62 out of 68 company trucks are operational per day. The whole fleet consumes around 4,292 gallons of diesel per day of operation, around 25,753 per week, and around 1,313,446 gallons of diesel per year. Each truck consumes around 21,184 gallons of diesel per year, 415 per week and 69 gallons per day of operation. The company has four 10"x80" underground diesel tanks which are capable of storing 180,000 gallons of diesel. Till now the tankers were being refueled every 4 weeks or at around 50% of tank filling level. The company would like to utilize the tanks more by adjusting their diesel ordering strategy by forecasting future diesel price change. Thanks to a combination of good forecasting model and experienced management which will make good use of it, a company would better utilize available resources and save money. For each cent saved on every gallon of diesel purchased, the company saves approximately 13,446 USD in a standard year of operation, which equals to around 211 USD per truck per year.

Note: The numbers above might not add up due to rounding.

Note about data

In order to get more information about the data, its collection and missing values, I highly recommend to visit the websites provided in the first section. The datasets downloaded from Bureau of Transportation Statistics might have missing values denoted as capital letters based on the reasoning behind the missingness. In the 'data note.txt' file I have included more information about BTS data. The datasets were downloaded as excel files and converted to csv files.

About

Truck transport US industry analysis and diesel price forecasting model using ML and traditional Data Science approach.

License:MIT License


Languages

Language:Jupyter Notebook 100.0%