论坛主题：Session 2: Big Transportation Data

时间：2017年05月11日（周四）14:50-15:50
地点：交通运输工程学院103会议室

主讲人1：Lijun Sun

主讲人照片：

主讲题目：Harnessing Data Analytics in Urban Public Transport Operation and Planning

主讲人简介

    Lijun Sun is a Postdoctoral Associate at MIT Media Lab (https://www.media.mit.edu/). His current research focuses on developing and applying crowdsourcing and data-driven approaches in the domain of civil systems and transportation. He earned his B.S. degree in Civil Engineering from Tsinghua University in 2011 and his Ph.D. degree in Transportation from National University of Singapore in 2015. Before joining MIT, he worked at Future Cities Laboratory (http://www.fcl.ethz.ch/), Singapore-ETH Centre as a PhD researcher and then a senior research fellow in the Mobility and Transportation Planning Module, combining smart card-driven public transport modeling and agent-based simulation to improve urban public transport service quality and reliability.

His research interests include data-driven transport modeling, resilience of urban transportation infrastructure and systems, mobility and travel behavior profiling, urban computing and complexity, and large-scale agent-based modeling/simulation. His research aims to provide a better understanding of urban and transportation systems and how scalable cooperation and artificial intelligence could benefit urban life. His work has been featured in popular media outlets, including Wired, Citylab, Scientific American and MIT Technology Review.

主讲内容简介

In recent years, the emergence of massive individual-based datasets and advances in informatics and data sciencehave transformed our understanding in a variety of fields in transportation research. It also motivates a new way of data-driven transport approach through conducting extensive analyses and building realistic models. For example, behavior pattern inference from spatial-temporal data set has facilitated the development of urban public transport in both day-to-day operation and long-term planning. However, in a large-scale and highly-congested city, the application of these technologiesremains prone to operational pitfalls and obstacles. In this talk, I will present about harnessing various data analytics and computational models to tackle resilience issues in public transport operation and planning, by integrating machine learning, operations research and behavioral economics.I will mainly discuss a typical topic about passenger behavior inference to show the basic concept of combining data analytics and computational models, and illustrate how it can be used to improve the resilience of public transport systems. The presented methodologies can be integrated with an agent-based modeling/simulation framework, and further help plan and evaluate future transportation systems (e.g., on-demand mobility services)for “resilient smart cities”.

主讲人2：Yanyan Xu

主讲人照片：

主讲题目：Big Data Driven Mobility to Tackle Urban Traffic and Electric Vehicle Management

主讲人简介：

Yanyan Xu is a postdoctoral associate in the Human Mobility and Networks Lab, Department of Civil and Environmental Engineering, MIT. He received PhD in Pattern Recognition and Intelligent Systems from Shanghai Jiao Tong University, China, in 2015. He works in the fields of data mining, human mobility, with a focus on the use of information and communication technology and big data in Transportation Systems, Environment, and Urban Planning. Specific projects include urban traffic flow prediction, travel demand management using big data; revealing the impacts of traffic on the air quality using mobile phone data; and coupling electric vehicle charging with urban mobility, etc. His work has been published in the J. Roy. Soc. Interface, IEEE Trans. ITS, J. Adv. Transp., TRB, among others.

主讲内容简介：

Understanding human mobility has many applications in diverse areas, including spread of diseases, city planning, traffic engineering, financial market forecasting, and nowcasting of economic well-being.In the past years, we have studied problems concerning the use of various sources of large-scale data to better inform human mobility and collective travel behavior in cities. In this presentation, I will introduce two main works on the applications of human mobility: traffic demand management and the Electric Vehicles charging planning. In the first task, our target is to understand the impact of mega events using multiple data resources and design feasible travel demand management strategy for a global mitigation of traffic congestion. In the second one, we couple the urban mobility with EV charging plan to alleviate the pressure of power grid from the EV charging. In detail, we represent the mobility of EV drivers in San Francisco Bay Area in USA, and shift the arrival and departure times of EV drivers to reduce the peak load of EV energy demand in the working area.

主讲人3：Weiliang Zeng

主讲人照片：

主讲题目：Exploring traveler’s risk preference to travel time uncertainty from large-scale GPS trip data

主讲人简介：
Dr. Weiliang Zeng is a postdoctoral researcher in Institution of Materials and Systems for Sustainability, Nagoya University, Japan. He received the B.S. and M.S. degrees from the College of Engineering, Sun Yat-sen University in 2009 and 2012, respectively. In 2012, he came to Nagoya University for further study, and received Phd degree in 2016.His current research interests include reliable routing problem, autonomous taxi system, pedestrian simulation etc. He received the best paper award in Eastern Asia Society for Transportation Studies (EASTS) in 2015. He is also the reviewer for TRC, TRE, IEEE ITS, Chinese Physics B，and Urban Rail Transit.

主讲内容简介：

Travel time reliability has found notable interest in route choice modelling. Knowing how individuals choose paths with uncertain travel times is fundamental to advance our understanding of route choice behavior and drive the development of route guidance systems. Current navigation systems usually provide shortest paths based on distance or travel time, despite many travelers do not choose the shortest one. Many studies pointed out that the probability of delay or travel time reliability is an important factor in traveler’s route choice decision. Traditionally, route choice data for individual preference analysis can be collected by stated preference surveys. However, it requires repeated interaction with the travelers, and inherent limitations related to honest, accurate and bias-free reporting are difficult to avoid. To fill this gap, this study proposes a new data collection methodology that enables to estimate traveler’s risk-averse preference by GPS trip record. The lower bound and upper bound of individual risk-averse preference can be estimated by exhausting a series of reliable paths with different on-time arrival probabilities and using the theory of stochastic dominance. Then, a regression model based on logistic function is established to explore how the influencing factors impact on the lower and upper bounds of risk-averse preference. It is found that the individual properties such as age and pre-trip information such OD distance, departure time, and day of week are significant to the degree of risk-averse preference.