DTUMOS
Digital Twin for Urban Mobility Operating System
Simulate urban mobility. Design better operations.
About
DTUMOS is an open-source digital twin for urban mobility operating systems. It is a virtual city for developing and testing mobility systems, algorithms and policies.
- 30,000+
- vehicles and 200,000 passengers handled in one metropolitan-scale system
- 100×
- faster travel-time estimates than direction APIs, from a learning-based ETA model
- 3
- metropolitan cities, Seoul, New York and Chicago, validated with real-world data
Data
Real-world records such as taxi trips, smart-card and mobile-signal data and delivery logs, plus the road network.
AI models
A learning-based ETA model corrects vehicle positions and travel times, for speed and accuracy at once.
Operation
A dispatch algorithm matches passengers and vehicles, which are routed on the road network minute by minute.
Outputs
Dashboards of waiting times, failed requests and fleet status, with smooth visualization of large fleets.
Team
Professor
여지호Jiho Yeo
Ph.D. program
김민수Minsu Kim
Master's program
정은주Eunjoo Jung
Master's program
방혜원Hyewon Bang
Master's program
송민경Minkyung Song
Master's program
장상준Sangjun Jang
Master's program
박준규Junkyu Park
Undergraduate researcher
김태우Taewoo Kim
Research
Paper2023
DTUMOS, digital twin for large-scale urban mobility operating systemScientific Reports, 13(1), 5154
Get in touch
Questions about DTUMOS, research collaboration or the class? Write to us.
- Lab & research
- jihoyeo@gachon.ac.kr
- DTUMOS class & login help
- junkyu@gachon.ac.kr
- Address
- Room 703, AI·Engineering Building, Gachon University, 1342 Seongnam-daero, Sujeong-gu, Seongnam-si, Gyeonggi-do, Korea
