The two-year Eurostars H2020 BAMBOO project is now complete. Aimsun has collaborated with Nommon Solutions and Technologies to develop new methods and algorithms for extracting travel demand information from anonymized mobile phone records and integrating the data into the Aimsun Next mobility modeling platform. The impressive visuals are courtesy of Ito World, who provided the interactive visualisation tools to bring it all to life.
The use cases were urban (Santander), interurban (the AP-7 highway) and national (Spain), each providing different data distribution and technical challenges, as well as different mobility objectives. Mobile-phone provider Orange provided access to large-scale tailored data and stakeholders for each use case.
The main focus of BAMBOO was to improve demand generation from mobile phone data, whether aggregated (OD trip matrices) or disaggregated (activity-travel diaries). Aimsun focused on overcoming the limitations associated with travel demand generation and representation in the context of dynamic traffic assignment, traditionally used in the sector to characterize demand derived on home-based surveys.
The new demand characterization offers richer and more reliable information for simulating more realistic traffic conditions.
Specifically, the Aimsun team worked on:
- Integrating new travel demand characterization from mobile phone data fused with other data sources into Aimsun Next software architecture;
- Developing new algorithms for identifying centers of activity and their connection to the road network;
- Improving the calibration and validation framework of a traffic network model using new demand characterization, including aspects such as value of time (VOT), trip departure time, mode and route choice.
The new features resulting from BAMBOO are now implemented in Aimsun Next software, which is now the first transport simulation toolset specifically adapted to realize the full potential of massive, passively collected data such as cellular data, as opposed to competing solutions that were conceived to work with traditional demand data. This is particularly important for modeling new forms of mobility and the most recent approaches to traffic management, for which the current tools typically fall short due to the need for higher levels of disaggregation and behavioral insights.