Adaptive Transportation and Logistics

Item 1 of 1

In Adaptive Transportation and Logistics (ADAPT-OR) group we bring together the methodologies of Operations Research, Behavioral Modeling and Machine Learning. Operations Research techniques are supported by Machine Learning algorithms to reach dynamic and predictive optimization models towards robust transportation systems. Behavioral models are incorporated into optimization models and demand-driven (aka choice-driven) decisions on the supply side are obtained where user preferences are incorporated. Machine Learning techniques also support the behavioral models as the preferences of users can be learned continuously as heterogenous users make choices and the overall system can make use of most up-to-date information. This integrated methodological framework enables to use the right resources at the right time that addresses the long term sustainability of the transport and logistics systems. A better utilization of the available resources is one of the key challenges towards sustainability goals and the methodologies we develop provide promising models and algorithms to address this.

News

ERC Starting Grant for developing adaptive transport systems with self-learning capabilities
[Translate to English:]

Modelling crowd behaviour for ultra-large ship design

Bilge Atasoy co-authored an open textbook: From theory to application

Representation of our group at conferences

Awards

Projects

Publications

About Bilge Atasoy

I am an associate professor in Transport Engineering and Logistics in the Department of Maritime and Transport Technology at TU Delft. I joined TU Delft in 2018 as an assistant professor and was promoted to Associate Professor in July 2022. I am working on Adaptive Transport and Logistics with my group with the core methodologies of Operations Research, Real-time Optimization, Dynamic and Predictive Optimization, Random Utility Models, Choice Models, Machine Learning, Bayesian Techniques.

Prior to joining TU Delft, I was a research scientist at MIT at the Intelligent Transportation Systems (ITS) Lab and was working with Prof. Moshe Ben-Akiva. At MIT, I managed research projects in the areas of real-time optimization, travel behavior and choice-based optimization. I am still a research affiliate of the ITS Lab. I obtained my PhD from EPFL in November 2013 with Prof. Michel Bierlaire at TRANSP-OR Lab. I obtained my MSc and BSc degrees in Industrial Engineering from Bogazici University, in 2009 and 2007, respectively.

B. (Bilge) Atasoy

We bring together Optimization, Behavioral Modeling and Machine Learning methodologies to reach Adaptive Transport and Logistics Systems.

The intersection between these methodologies enable to have (1) a predictive decision making towards robustnesss (2) a user-centric nature for better incorporation of preferences (3) an adaptive system that continuously evolves towards better decisions and better representation of user behavior.

We holistically represent transportation systems with the supply and demand sides as well as the interactions between them. Different decision making levels on each side are modeled in such a way that they interact with the aim of reaching better and more representative decisions at each level. The system is continuously fed with historical and real-time data and adapted as necessary.

People

B. Atasoy (Bilge)

C. Karademir (Cigdem)

A. Nicolet (Adrien)

P. Zattoni Scroccaro (Pedro)

V.M.J. Ramos (Jayvee)

Karel Scheepstra

J. Duran Micco (Javier)

  • Yimeng Zhang (PhD graduated in June 2023). Now postdoc at Transport and Planning at CITG TU Delft. 
  • Breno A. Beirigo (Postdoc finished in September 2022) Now assistant professor at University of Twente 
  • Rie Larsen (PhD graduated in September 2022) Now postdoc in our department 
  • Jishnu Narayan (Postdoc finished in Jan 2022) Now Transport Planner at Atkins 
  • Wenjing Guo (PhD graduated in Oct 2020) Now assistant professor at Wuhan university of Technology