Smartphone-based solutions to monitor and reduce fuel consumption and CO2 footprint : final report.

Details

• Creators:
  Cetin, Mecit ; Ustun, Ilyas ; Nadeem, Tamer ; Nguyen, Duc ; Rakha, Hesham A.
• Corporate Creators:
  Old Dominion University. Transportation Research Institute
• Corporate Contributors:
  United States. Department of Transportation. University Transportation Centers (UTC) Program
• Subject/TRT Terms:
  Carbon Dioxide Machine Learning CO2 Environmental Impacts Fuel Conservation Fuel Consumption Greenhouse Gases Mobile Applications Mode Choice Smartphones Travel Time
• Publication/ Report Number:
  N16-01
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States
• Edition:
  Final report, Jan 1, 2014 - Jan 31, 2016
• Corporate Publisher:
  TranLIVE. University of Idaho
• Abstract:
  Smartphones equipped with GPS and several low-energy sensors (e.g., gyroscope, compass, and accelerometer) can provide
  
  a medium to collect probe data. As smartphone users navigate the transportation networks, their travel modes and
  
  trajectories can be inferred to estimate fuel consumption and CO2 footprint. The specific goals of the proposed research are:
  
  (i) Develop new algorithms to estimate the mode of travel (walking, biking, train, car, bus, etc.) and operating mode of a
  
  vehicle (e.g., idling) based on low-energy sensors available within smartphones; (2) Evaluate the effectiveness of FC and
  
  CO2 estimation from probe vehicles at various market penetration levels; and (3) Develop shortest paths algorithms for
  
  finding eco-friendly routes. To achieve these goals, various methodologies are developed and tested with both simulation
  
  and field data. For example, machine learning algorithms (e.g., support vector machines) are developed to predict the travel
  
  mode and to detect whether a vehicle has stopped. The results show that the travel mode can be detected accurately, about
  
  94% on average, when considering all five travel modes within the sample data. Using the accelerometer data only, the
  
  results show that the models can accurately detect the times at which the vehicle stops and moves during its journey. To
  
  predict the impacts of probe vehicle market penetration on estimating fuel consumption, simulation data are created for an
  
  intersection. Lastly, the shortest path (SP) algorithm for static networks are modified so that the algorithm can find the SP
  
  for minimizing both the travel time and fuel consumption for a given network. Together, the models and algorithms
  
  developed in this study can be integrated to support various mobility and environmental applications.
  
  
• Format:
  PDF
• Funding:
  DTRT12GUTC17
• Collection(s):
  University Transportation Centers US Transportation Collection
• Main Document Checksum:
  urn:sha-512:3a786a1759e8df3b42c11a77bf516f271422adab67c1d6cdb2eea6234ce972cf06f4022e8857d32f8dc57df7e496a89aadaf2f8db7ababacd4d9f2de25b465a2
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/32549/dot_32549_DS1.pdf
• File Type:
  [PDF - 798.03 KB ]