Interfaces Between Optimization and Practice
Optimizing Freight Transport by representing uncertainty, behavior and cooperation
Freight transport is going through various developments towards increasing the adaptive nature of the several decisions involved. One of those developments is the concept of synchromodality where real-time planning is emphasized. In this talk, I will talk about key methodological developments that enable to achieve the ideas behind such innovative concepts. First, optimization of operational decisions under uncertainty is presented using stochastic optimization as well as reinforcement learning. Then, incorporation of behavioral elements for operational decisions of synchromodal transport are presented. Finally, I will bring the attention to cooperation to facilitate more efficient and effective real-time planning models.
Duty Rostering for Physicians: A Web-Based System and its Practical Application
Duty rostering for nurses and physicians is an important task within personnel planning in hospitals and has a large impact both on the efficient operation of the hospital and on employee satisfaction. Good rosters should not only satisfy many complex constraints resulting, e.g., from minimum rest times or required staffing levels, but at the same time achieve a fair distribution of the workload and adhere to the preferences of the planned personnel.
This talk presents a web-based duty rostering system for physicians that is used in practice in two different departments of two German hospitals. The system consists of custom-built integer-programming-based optimization models for duty roster generation and a web interface that is used to collect all necessary input data such as each physician's preferences concerning each possible duty assignment on each day of the planning period. Besides the structure of the duty rostering problems faced in the two departments, we present how fairness and physician preferences are incorporated and we demonstrate how the physicians' preferences are elicited via the web interface. Moreover, we analyze the computed rosters and present long-term results about their quality.
Revenue Management with Two Fare Classes in Synchromodal Container Transportation
Rommert Dekker, Bart van Riessen, Judith Mulder and Rudy Negenborn
We consider two fare classes for intermodal transport on a network. The problem is to select two fare class acceptance limits on a tactical level up to which transportation demand will be accepted on a daily basis at the operational level. Demand which cannot be accepted has to be transported in a more costly way. We first show the method on a single corridor and then extend it to multiple corridors. Finally we apply the methods in a case study for an intermodal transportation network in NW Europe.
Online Zoom Details:
Meeting URL: eur-nl.zoom.us/j/95792528123
Meeting ID: 957 9252 8123