The SORTEDMOBILITY team is happy to invite you to the project final event that will take place in Paris on Monday May 13th, 2024, 10am to 5pm.
SNCF will host the event at 34 Rue du Commandant René Mouchotte, 75014 Paris.
The agenda for the day includes crash courses on cutting-edge research topics and the presentation of the project results. Ample room for discussions and networking will also be granted (see the detailed schedule below).

Registration is free but compulsory for organizational reasons (expired).

We hope to see you there to enjoy some exciting exchanges on the future of railway traffic management.

The SORTEDMOBILITY team

Agenda

09:30 - 10:00     Welcome with coffee
10:00 - 10:10     Introduction
10:10 - 11:10     Crash course for practitioners: Machine learning for planning and decision making
11:10 - 11:30     Coffee break
11:30 - 12:15     Crash course for practitioners: Collective intelligence for decision support
12:15 - 12:30     SORTEDMOBILITY: paving the way to decentralized traffic management
12:30 - 13:50     Buffet lunch
        SORTEDMOBILITY:
13:50 - 14:10     Train self-organization for traffic management decisions
14:10 - 14:30     Consideration of dynamic demand requirements for traffic management
14:30 - 14:40     Joint simulation of rail passenger and operations
14:40 - 15:10     Evaluation of self-organizing traffic management in different case studies
15:10 - 15:30     Recommendations
15:30 - 16:00     Coffee break
16:00 - 17:00     Round table: next steps
17:00 – 17:10     Conclusions

Crash courses for practitioners
Machine learning for planning and decision making
The technological changes (such as automation electrification) and the global challenges (e.g., social justice and climate change) faced by the mobility agenda are disrupting the sector, and AI can play a central role in the planning and decision-making of future mobility systems. Traditional modeling approaches cannot often take the most out of large datasets and the computational advantages of AI. Data-driven approaches to decision-making can help bridge the gap and provide flexible, sustainable tools that can support and supplement transport planning and operation efforts. We look at fundamental concepts in machine learning and dive into the concepts of Bayesian Optimization and Reinforcement Learning, looking at specific applications in transportation.

Collective intelligence for decision support
Human knowledge is growing exponentially, providing huge and sometimes contrasting evidence to support decision making in the realm of complex problems. A promising way to improve decision making is exploiting collective intelligence, which integrates the advice of multiple individuals and provides decision support. While the power of collective intelligence has been successfully demonstrated in multiple domains, it was often applied to rather simple decision problem. This course starts from collective decision-making problems that can be solved through consensus processes inspired to self-organizing biological systems such as honeybee swarms. Then, we move to settings in which humans and artificial systems collaborate to address complex decision problems. We discuss how such a hybrid human-artificial collective intelligence can be deployed, and we present case studies in different domains.