Analyzing Public Transportation Offer wrt Mobility Demand

Abstract

An important problem in improving mobility services consists in analyzing the transportation offer with respect to the demand of mobility. The purpose is always the assessment of the service for its improvements. This activity can be approached having all the historical data, while in most cases is not realistic due to the expensive process of data collection and lack of details about the movements of travelers at the bus stops in terms of pickup and drop-off for each bus line. To deal with these issues, in this paper, a model is provided to support mobility analysis in public transport networks. Our model operates first by analyzing the service offer, provided by mobility operators, and the service demands. Then, the model allows to evaluate the number of people who are picked-up and dropped-off at a stop. The performance of the model has been validated by comparing the observed values obtained from a field observation. The research and tool have been developed in the context of MOSAiC research project partially funded by Tuscany Region, with DISIT lab, ALSTOM, Municipia/Engineering, TAGES and CNIT research centers.

Automating IoT Data Ingestion Enabling Visual Representation

“…The authors in [26] proposed universal meta-modeling for the big data storage layer to address the problem of the diversity of existing solutions by presenting a shift from a generic big data storage layer meta-model, based on Hadoop [27], to a Cloudera [28] distribution storage layer. This was in contrast to a considerable amount of smart city solutions supporting only a single domain (e.g., air pollution monitoring [29], roadside assistance [30], smart parking [31], urban [32] and public [33] transportation, smart power grids [24], and smart meters [34,35]). Therefore, when it comes to a smart city solution it must be possible to ingest, store, and process any heterogeneous multi-dimensional data from different domains, with the possibility of defining new ones.…”

Self-Attention based encoder-Decoder for multistep human density prediction

“…Following, these data should be carefully analyzed to extract the appropriate knowledge that will make relevant applications more efficient and more intelligent at the same time. To this end, the key theme of this paper is to design a data driven and machine learning model that processes the mobility data in order to provide timely and accurate insight for an optimal decision making and what-if ( Arman et al, 2019 ) analysis in the context of urban mobility. Specifically, the problem we address is the Multistep Human Density Prediction (MHDP).…”