Optimization of PMPML Bus Routes

• A pilot project was implemented for Pune Mahanagar Parivahan Mahamandal Limited (PMPML) to analyze the existing 47 city bus routes in terms of revenue and passenger traffic and to suggest new routes, based on the inferences from statistical analysis.
• The latitude and longitude of different stops were analyzed using the base map Python package. The 47 test routes appear in Figure.

These routes are further analyzed to identify the extent of overlap of routes. It is identified that there are 11 sets of routes with more than 80 % overlap.

• The timewise and stagewise analysis of the passenger count in these routes are done to analyze the feasibility of different trips in these routes.
• Based on these analyses different suggestions were given to the PMPML authorities, which were implemented there, resulting in a rise in epk (earning per kilometer) and showing marked improvement in performance.

• GPR is a supervised machine learning technique that provides a mapping from input to output.
• The training procedure adopted by the GPR process is that it considers the whole training data set at each time the model makes a prediction.
• The GPR-based models can easily overcome overlearning or other data-driven issues since the training process considers the whole set of training data. Thus GPR is selected for the prediction of passenger traffic.
• It is required to maintain a uniform number of passengers in all stages to increase passenger comfort and revenue. The arrival of passengers is conjectured as a sum of several time-varying Poisson processes.
• The passenger traffic at any instant shows random behavior and does not depend on the previous passenger history, making it a Markovian process.
• The methodology and experimental steps are as shown