28. Februar – 04. März 2016, Dagstuhl Seminar 16091
Computational Challenges in Cooperative Intelligent Urban Transport
1 / 3 >
Auskunft zu diesem Dagstuhl Seminar erteilt
Following the history of two Dagstuhl seminars on Computational Issues in Transportation in 2010 and 2013, the organizers of this follow-up seminar concentrated on upcoming, data-driven challenges in the area of urban transport. In recent years, urban transportation networks have become more diverse, with a growing mix of public and private operators providing disaggregated services and information. The resulting multitude of transportation options includes non-traditional modes and services such as car and bike sharing in addition to established public transport and individual car options. So far, it is challenging to combine detailed operational data automatically arising from these services, since these data are generated both from service operation and from the users of services via crowdsourcing. The seminar aimed to discuss how data sources can be made available for individual planning and system-wide coordination of urban transportation using an approach from distributed computing, i.e., getting all involved parties to cooperate in providing relevant spatial and temporal information in a timely fashion. It was not clear how to derive reliable information for planning and control approaches, or how to adapt optimization methodologies to make urban transportation more cooperative and intelligent.
The aims of the seminar were to extend the existing network in disciplines such as Computational Traffic Science, Optimization, Autonomic Computing and Artificial Intelligence for discussing computational challenges in cooperative intelligent urban transportation, mesh communities by collecting suggestions for (partial) solutions for burning issues in urban transportation and discussing the prerequisites for merging into interdisciplinary approaches, document the state of the art and current computational challenges in cooperative intelligent transportation.
To this end, an interdisciplinary group from areas such as computer science, geography, applied optimization and traffic engineering met at Dagstuhl. The number of attendees was advantageous for group discussions, not too small for breakout groups but also not too large for meaningful discussions in the plenum.
We started on Sunday evening with a game ("Cards Against Urbanity -- special issue for this seminar") specifically designed for this event by Ms. Cottrill. The game was a great success as icebreaker and helped bringing together the participants with their various backgrounds. Monday was opened with a keynote by Vonu Thakuriah, who discussed examples, prospects and challenges of emerging forms of data in transportation research and applications. The participants introduced themselves, bringing a significant object describing their relationship with the seminar’s topic.
For the remaining seminar time, the participants were asked to contribute to the seminar’s content by one of the following options: they could give an overview talk of an emerging area (20 minutes), a research statement on what they have been working on in their particular area (5 minutes), and they were asked to come together in groups that were defined dynamically on Monday afternoon. The resulting abstracts can be found in this report. Based on the participants’ interests, groups discussing the topics of online simulation, pedestrian behavior, autonomous transportation, smart cities, and benchmark data emerged. On Wednesday afternoon, the participants went on a ‘field trip’ to the retail lab by DFKI in St. Wendel, where the future of retail can be explored hands-on. Since there was a significant interest in the provision of benchmark data for urban transport, there was a special session and group work devoted to this topic on Thursday afternoon. Friday morning was meant for collecting the results of the group work and collecting open challenges for future seminars.
Summarizing, the seminar identified computational challenges to cooperative intelligent urban transport, among others notably research on opportunistic groups in public transport (i.e., people sharing tickets and or trajectories in an ad-hoc fashion), freight pods attached to light rail (i.e., mixing of freight and passenger transportation), define a common language for sharing complex knowledge and real-time data in smart cities and creating benchmark datasets for different modelling purposes and at different scales. We think that the seminar was quite successful in extending the existing networks by bringing together researchers from many different disciplines relevant for the future of urban transport. Some of the groups are planning to write proposals for the appropriate EU calls coming out in October, while others have started to work on position papers describing the state of the art as well as resulting future challenges of the field.
Creative Commons BY 3.0 Unported license
Caitlin Doyle Cottrill and Jan Fabian Ehmke and Franziska Klügl and Sabine Timpf
Dagstuhl Seminar Series
- 13512: "Social Issues in Computational Transportation Science" (2013)
- 10121: "Computational Transportation Science" (2010)
- Mobile Computing
- Modelling / Simulation
- Society / Human-computer Interaction
- Computational transportation science
- Intelligent transportation systems
- Cooperative computing
- Crowd-sourcing of transportation data