Passenger occupancy prediction in public road transport

Description 

The inLab FIB UPC collaborates with Intelibus in a project to characterize and model passenger demand for buses.

The main objective is the development of algorithms for predicting passenger occupancy on a bus using data-driven methodologies using data from heterogeneous sources (ticketing, calendar, cameras on buses, etc.). In particular, classical methods for the treatment of time series will be studied, such as the ARIMA method, and neural network algorithms that will allow other important variables such as the school calendar to be incorporated into the occupancy time series.
 
Intelibus is a real-time information system in the passenger transport sector, which uses the GPS positions of buses and the ticketing or counting system, to provide useful and real-time information to both public transport users and service operating companies.

The results of the investigation will allow the user to consult, through an APP, the estimated occupations on buses throughout the day. In this way, the aim is to ensure that users opt for those expeditions that are the most empty, thus balancing the supply and demand for the service, and avoid crowds on public transport.

The new tool will be launched by the transport operator company Autocorb, which operates urban and interurban bus lines in the Barcelona metropolitan area.

Duration of the project 
January, 2020 - October, 2020
Client 
Funded by 
Technology 
Python,
Scikit,
Keras,
Tensorflow,
Numpy,
Pandas,
Matplotlib,
Google Colab Notebooks,
Google BigQuery
Areas of expertise involved in the project 
Project Manager 

Segueix-nos a

Els nostres articles del bloc d'inLab FIB

         
         

inLab FIB incorporates esCert

Icona ESCERT

First LogoCSIRT Logo

inLab is member of

inLab és centre TECNIO

ACCIO