BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:ENAC Seminar Series by Dr K. Wang
DTSTART:20210412T140000
DTEND:20210412T150000
DTSTAMP:20260511T105817Z
UID:67edc90f63b73726f25b7da2665f1222a91085cc95591c9d37ed5cbc
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr Kai Wang\n14:00 – 15:00 – Dr K. Wang\nPostdoctoral Asso
 ciate at Massachusetts Institute of Technology\, USA\n\nVertiport Planning
  for Urban Aerial Mobility: An Adaptive Discretization Approach\n\nMobilit
 y systems are experiencing unprecedented transformations due to rapid tech
 nological developments (e.g.\, vehicle technologies\, information and comm
 unication technologies) and new business models (e.g.\, ride-sharing\, on-
 demand business). At the same time\, the mobility sector is facing increas
 ing pressure to meet sustainability goals\, ranging from mitigating climat
 e change to creating equitable access. Addressing these challenges require
 s novel analytics and optimization methods to design a pathway toward huma
 n-centric mobility. This talk explores these broad themes\, with a particu
 lar focus on Urban Aerial Mobility (UAM)\, enabled by the rise of electric
  vertical-takeoff-and-landing (eVTOL) vehicles.\n\nThe immediate decision 
 faced by the future UAM operator involves determining the number\, locatio
 n\, and capacity of vertiports in a metropolitan area. We formulate an opt
 imization model that captures interdependencies between strategic vertipor
 t deployment\, tactical eVTOL operations\, and customer adoption. The mode
 l includes a “tractable part” (a mixed-integer second-order conic prog
 ram) but also a non-convex customer adoption function that makes it intrac
 table—a common problem when capturing demand-supply interactions in mobi
 lity systems. We develop an original exact algorithm based on adaptive dis
 cretization\, which dynamically refines “promising” regions while reta
 ining computational tractability. Computational results suggest that the a
 lgorithm converges to a 1% optimality gap\, dominating traditional static 
 discretization in terms of solution quality\, runtimes\, and solution guar
 antee. From a practical standpoint\, we find that UAM networks vary widely
  across metropolitan areas\, as a function of geographic\, urban\, and com
 muting patterns. Vertiport networks grow in a nested fashion\, starting wi
 th a few “obvious” vertiports and adding vertiports as penetration inc
 reases. Operationally\, UAM supports primarily long trips\, with complex d
 ynamics to enable passenger pooling and vehicle recharging. We uncover two
  potential use cases for UAM technologies: airport shuttle and long-distan
 ce commutes.\n\n\nShort bio:\nKai Wang is a Postdoctoral Associate from th
 e Massachusetts Institute of Technology. He obtained his PhD degree from T
 he Hong Kong Polytechnic University in 2019. He was also a visiting PhD st
 udent at Carnegie Mellon University. Kai Wang’s research spans large-sca
 le\, stochastic\, and data-driven optimization\, with applications in mobi
 lity and logistics systems. His research has tackled a wide range of real-
 world problems\, spanning urban mobility\, aviation\, maritime transportat
 ion\, and smart cities. His research has appeared in top-tier journals suc
 h as Operations Research\, Transportation Science\, and Transportation Res
 earch Part B. It has been recognized by several academic distinctions\, e.
 g.\, the Best Paper Award in the Applied Track from the 15th INFORMS Works
 hop on Data Mining and Decision Analytics (2020).\n 
LOCATION:https://epfl.zoom.us/j/89146322284
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR