Informs Annual Meeting 2017

TE13

INFORMS Houston – 2017

2 - On-demand Public Bus Service on a Multimodal Travel Corridor Jie Zhang, PhD Student, Nanyang Technological University, N1-B4-B06, 50 Nanyang Avenue, Singapore 639798, Singapore, 639798, Singapore, jzhang049@e.ntu.edu.sg, Zhiwei Wang On-demand Public Bus (ODPB) service, a new type of public transport service, has emerged in many big cities. This study analyzes the provision of ODPB by applying analytical approaches based on demand analysis and network modeling. Several alternative travel modes are considered. The choices of mode and route/service are taken into account simultaneously in logit-type formula. The resultant traffic equilibrium condition is expressed as a variational inequality (VI) problem. Numerical analysis from one idealized corridor indicates that the introduction of ODPB service could effectively bring a significant impact on mode shift, especially for the long distance trip. 3 - Design Optimal Strategy of On-demand Mobility Allowance Shuttle Transit (MAST) Service Dahye Lee, Texas A&M.University, 3135 TAMU, College Station, TX, 77843-3135, United States, dahyelee1991@tamu.edu, Luca Quadrifoglio, Kai Yin We adopt the Markov Decision Process (MDP) with empirical dynamic programming algorithms to design optimal strategies for Mobility Allowance Shuttle Transit (MAST) service with demand uncertainties. The MAST allows vehicles to deviate from fixed routes to serve customers at their desired pick-up and drop-off locations within the service area. The main objective of this study is to then obtain the optimal policy of vehicle routing strategy that minimizes the total cost subject to the vehicle distance traveled and the number of served customers, and evaluate the quality of solutions with numerical studies. 4 - Bus Fleet Turnover Strategies to Optimize Life-cycle Costs and Carbon Emission Reductions Nicholas E. Lownes, University of Connecticut, 261 Glenbrook Road, Unit 3037, Storrs, CT, 06269-3037, United States, nicholas.lownes@uconn.edu, Adrita Islam The increasing competitiveness of battery electric bus (BEB) technology is prompting transit agencies across the globe to consider a transition to a BEB fleet. In most cases, this transition will occur over a number of years, depending on the age composition of the bus fleet. BEB stand to generate substantial mobile GHG savings and are seen as increasingly cost competitive with standard diesel hybrid buses, with cost parity anticipated to occur sometime within the next decade. This paper presents a fleet turnover optimization model that considers life-cycle costs and potential GHG reductions. An application on the Connecticut bus fleet is presented to illustrate results of the model. 332C Decision Support Systems Contributed Session Chair: Majed Al-Shawa, Strategic Actions, Waterloo, ON, Canada, malshawa@strategic-actions.com 1 - An Integrated Feedback and Optimization Approach for Healthcare Services This work addresses an integrated service quality and multi objective optimization problem in the context of healthcare delivery. Feedback on the experience in the hospitals concerning the clinical and process parameters have been collected. Machine learning techniques have been applied to recommend a healthcare delivery. Moreover, we feel that an integrated approach involving feedback and healthcare operational process parameters more scientifically reflect the quality of care. Hence an optimization problem involving the operational parameters has been studied using evolutionary intelligence. 2 - An Analysis of Pricing Models Between Online Medical Service and Traditional Hospital Donghui Yang, Southeast University, Sipailou 2, Nanjing, 210096, China, dhyang@seu.edu.cn With new developments related to online technologies and environments, medical online is much more a feasible venue for physicians and patients. In this work, three scenarios (online consulting, see a doctor online or in a hospital, see a doctor online or in a hospital with drug purchasing option) are compared to help a patient to make a good decision. Meanwhile, a game between online medical and a hospital is constructed to optimize their profits. TE13 Soumyajyoti Datta, Indian Institute of Management Indore, Indore, India, f14soumyajyotid@iimidr.ac.in, Rohit Kapoor

3 - Approximation of Max Min Fairness of Bandwidth Allocation in Multicommodity Flow Networks Hamoud S. Bin Obaid, PhD candidate, University of Oklahoma, 1526 E.Lindsey Street, Apt I, Norman, OK, 73071, United States, hsbinobaid@ou.edu, Theodore B.Trafalis A bi-objective LP model to find the max min fair flow vector in multicommodity flow networks is proposed. The overall throughput is maximized in the first objective, and the sum of absolute deviations among commodities is minimized in the second objective. The model is capable of solving large-scale networks with low computational time producing a quality solution. The model is tested on real and random networks. 4 - RISKMETER: An Interactive Tool for Integrating Public Preferences into Long-range Electricity Supply Planning Georgios Xexakis, ETH Zurich, Universitatstrasse 22, K76.1, Zurich, CH-8092, Switzerland, georgios.xexakis@usys.ethz.ch, Sandra Volken, Evelina Trutnevyte Finding consensus pathways for transforming national electricity supply systems is a challenge due to multi-dimensional policy priorities and complexity of public preferences. We use a linear optimization model EXPANSE with Modeling to Generate Alternatives (MGA) to quantify 500 electricity supply scenarios for Switzerland in 2035. We then assess the scenarios in terms of health, safety, built and natural environment impacts and use an interactive tool RISKMETER to visualize the impact tradeoffs. Ultimately, we conduct an online survey (N=200) with non-experts to empirically evaluate the RISKMETER’s usability, collect public preferences, and integrate these with the EXPANSE scenarios. 5 - Modelling and Analyzing Stabilities and Equilibrium Strengths of Strategic Non-cooperative Conflicts using the Constrained Rationality Framework Majed Al-Shawa, Strategic Actions, Waterloo, ON, N2V 2V8, Canada, malshawa@strategic-actions.com We discuss the theory of moves and counter-moves of decision making agents that Constrained Rationality proposes for strategic multi-agent non-cooperative conflicts, and define four different stability and equilibrium solution concepts to analyze such conflicts. These concepts will guide the stability analysis of the conflict’s states, for each agent. Then, we define the stability strength for each solution concept and the strength of the equilibrium that could result from it. To demonstrate the effectiveness of the framework, we model and analyze the Cuban Missile Crisis, and show why the sequence of events observed in reality was the most rational sequence of events for the conflict to take. 332D Modeling to Improve Healthcare System Performance Sponsored: Manufacturing & Service Oper Mgmt, Healthcare Operations Sponsored Session Chair: Bryan A. Norman, University of Pittsburgh, University of Pittsburgh, Pittsburgh, PA, 15261, United States, banorman@pitt.edu 1 - A Technical Note on (r,s,q) Inventory Control Systems in Hospitals Bryan A. Norman, University of Pittsburgh, Department of Industrial Engineering, 1006 Benedum Hall, Pittsburgh, PA, 15261, United States, banorman@pitt.edu, Nazanin Esmaili, Jayant Rajgopal We develop discrete time Markov chain models of (R,s,Q) inventory control systems which are used at Point of Use Locations (POU) in hospitals and characterized by limited storage capacity, stochastic demand, periodic reviews with fractional lead times, expedited delivery when stocking out, and very high service level requirements. We present a novel approach that locally decomposes the state space for this specific class of inventory systems and use this to reduce the number of equations that must be solved by at least a factor of two. The result is that the overall computational effort required for finding the optimal replenishment policy parameters is significantly reduced. 2 - Wait Time Announcements at Hospital Emergency Departments Marco Bijvank, University of Calgary, 2500 University Dr. NW, Calgary, AB, T2N.1N4, Canada, marco.bijvank@haskayne.ucalgary.ca, Zhankun Sun A number of Canadian hospitals have started publishing live emergency department (ED) wait times online in an effort to provide patients with expectations on how long they will have to wait to be seen for non-urgent care after initial assessment by a triage nurse. We accurately predict the state- dependent wait times at emergency departments based on a busy-period analysis for a multi-class, multi-server priority queue with delayed feedback. We illustrate the robustness and impact of the predictor on patient flow and patient care with a case study at four major hospitals in the Calgary area. TE14

414