PhD Defence Pieter van den Berg (Stochastics)

Everyone is welcome to attend the public defense of Pieter van de Berg of his thesis                 'Logistics of emergency response vehicles: facility location, routing, and shift scheduling' Promotors: Prof. dr. ir. K.I. Aardal en prof. dr. R.D. van der Mei
  • PhD Defence Pieter van den Berg (Stochastics)
  • 2016-06-06T12:30:00+02:00
  • 2016-06-06T14:00:00+02:00
  • Everyone is welcome to attend the public defense of Pieter van de Berg of his thesis                 'Logistics of emergency response vehicles: facility location, routing, and shift scheduling' Promotors: Prof. dr. ir. K.I. Aardal en prof. dr. R.D. van der Mei
  • What Stochastics
  • When Jun 06, 2016 from 12:30 PM to 02:00 PM (Europe/Amsterdam / UTC200)
  • Where Senaatszaal of the Auditorium of the TU Delft, Mekelweg 5, Delft
  • Add event to calendar iCal

Everyone is welcome to attend the public defense of Pieter van de Berg of his thesis                 'Logistics of emergency response vehicles: facility location, routing, and shift scheduling'

Promotors: Prof. dr. ir. K.I. Aardal en prof. dr. R.D. van der Mei

At 13:00 committee member prof. Shane Henderson, from the Cornell University, will give a lecture on his work for the Bike-sharing service in New York, Citibike. This lecture will take place in room F of the EWI building of the TU Delft (Mekelweg 4, Delft)

(Citi)bike Sharing

Citibike has provided a bike-sharing service in New York City since March of 2013. We have assisted with their logistics since that time. The central decision in New York is one that is common to all bike-sharing programs, namely where bikes and racks should be positioned around the city. I will discuss a stochastic model of bike flows and an optimization model that uses the results from the stochastic model. This optimization model is a nonlinear integer program in approximately 700 variables. Geometric and coupling arguments reduce the problem to a linear (integer) program, which we can solve to optimality. The resulting plans yield substantial improvements in performance, and Citibike has adopted this work in its planning. I will also describe a simulation-optimization approach that determines a bike allocation assuming a fixed rack allocation, and that differs substantially from the usual simulation-optimization approaches used by the simulation community. Perhaps surprisingly, the simulation optimization yields much more than modest additional improvements over and above those obtained above.

 

Joint work with Ph.D. students Daniel Freund, Nanjing Jian, Eoin O'Mahony, several undergraduate researchers, and David Shmoys.

 

See also the news item (in Dutch): 'Wiskundige modellen voor optimale dekkingsgraad hulpdiensten'