Optimal Green Fleet Composition Using Machine Learning
Due to the use of petroleum-based fuel, the transportation sector is one of the two principal contributors to greenhouse gas emissions and its contributions are expected to double by 2050. Freight sector contributes to around 30% of all transport related CO2 emissions. Since different type of vehicl...
| Main Authors: | , |
|---|---|
| Published: |
2019
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/1721.1/121322 |
| _version_ | 1826203996108357632 |
|---|---|
| author | Patil, Vrushali Samaha, Elissar |
| author_facet | Patil, Vrushali Samaha, Elissar |
| author_sort | Patil, Vrushali |
| collection | MIT |
| description | Due to the use of petroleum-based fuel, the transportation sector is one of the two principal contributors to greenhouse gas emissions and its contributions are expected to double by 2050. Freight sector contributes to around 30% of all transport related CO2 emissions. Since different type of vehicles exhibit different fuel efficiency when operating in different regions and under different load conditions, companies face the challenge of determining which vehicles are more fuel-efficient and have better emissions performance. In this study, we asses carbon emissions and fuel efficiency characteristics of delivery trucks in the inbound delivery fleet for one of the largest retail companies in Mexico: Coppel. Coppel’s inbound fleet consists of 590 trucks, operating in diverse geographies throughout Mexico, making it difficult to direct compare their fuel efficiency. We use machine learning algorithms to analyze Coppel’s trucks’ performance and examined their fuel efficiency for varying road and different traffic conditions. We use these insights to build a green fleet optimization model that considers costs and CO2 emissions performance. By running different scenarios, we observe solutions where CO2 emissions drop by 3.5% with 0.04% increase in costs for Coppel’s inbound fleet. We also observe evidence that brand and age play an important role in the CO2 emissions performance of the vehicles. |
| first_indexed | 2024-09-23T12:47:11Z |
| id | mit-1721.1/121322 |
| institution | Massachusetts Institute of Technology |
| last_indexed | 2024-09-23T12:47:11Z |
| publishDate | 2019 |
| record_format | dspace |
| spelling | mit-1721.1/1213222019-06-19T03:01:07Z Optimal Green Fleet Composition Using Machine Learning Patil, Vrushali Samaha, Elissar Optimization Strategy Environment Urban Logistics Due to the use of petroleum-based fuel, the transportation sector is one of the two principal contributors to greenhouse gas emissions and its contributions are expected to double by 2050. Freight sector contributes to around 30% of all transport related CO2 emissions. Since different type of vehicles exhibit different fuel efficiency when operating in different regions and under different load conditions, companies face the challenge of determining which vehicles are more fuel-efficient and have better emissions performance. In this study, we asses carbon emissions and fuel efficiency characteristics of delivery trucks in the inbound delivery fleet for one of the largest retail companies in Mexico: Coppel. Coppel’s inbound fleet consists of 590 trucks, operating in diverse geographies throughout Mexico, making it difficult to direct compare their fuel efficiency. We use machine learning algorithms to analyze Coppel’s trucks’ performance and examined their fuel efficiency for varying road and different traffic conditions. We use these insights to build a green fleet optimization model that considers costs and CO2 emissions performance. By running different scenarios, we observe solutions where CO2 emissions drop by 3.5% with 0.04% increase in costs for Coppel’s inbound fleet. We also observe evidence that brand and age play an important role in the CO2 emissions performance of the vehicles. 2019-06-17T15:40:54Z 2019-06-17T15:40:54Z 2019 https://hdl.handle.net/1721.1/121322 application/pdf |
| spellingShingle | Optimization Strategy Environment Urban Logistics Patil, Vrushali Samaha, Elissar Optimal Green Fleet Composition Using Machine Learning |
| title | Optimal Green Fleet Composition Using Machine Learning |
| title_full | Optimal Green Fleet Composition Using Machine Learning |
| title_fullStr | Optimal Green Fleet Composition Using Machine Learning |
| title_full_unstemmed | Optimal Green Fleet Composition Using Machine Learning |
| title_short | Optimal Green Fleet Composition Using Machine Learning |
| title_sort | optimal green fleet composition using machine learning |
| topic | Optimization Strategy Environment Urban Logistics |
| url | https://hdl.handle.net/1721.1/121322 |
| work_keys_str_mv | AT patilvrushali optimalgreenfleetcompositionusingmachinelearning AT samahaelissar optimalgreenfleetcompositionusingmachinelearning |