Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MD40

5 - Disruption Risk Management in Serial Multi-echelon Supply Chains Florian Lucker, Cass Business School, 106 Bunhill Row, London, EC1Y 8TZ, United Kingdom, Sunil Chopra This talk deals with managing supply chain disruption risk in a multi-echelon serial supply chain using risk mitigation inventory and reverse capacity. We show that it is typically optimal to hold more risk mitigation inventory downstream than upstream. At the same time, it is often optimal to hold additionally more reserve capacity downstream than upstream. These results hold under the assumption that inventory and reserve capacity holding costs are larger downstream than upstream. n MD40 North Bldg 226B Utility Theory Developments Sponsored: Decision Analysis Sponsored Session Chair: Ying He 1 - Shapley Values, Interactions and Supermodularity Emanuele Borgonovo, Bocconi University, Via Sarfatti 25, Milano, 20136, Italy, Giovanni Rabitti In this work, we present a miscellanea of results concerning the study of interactions in decision analysis modelling. We show the link between interactions and supermodularity. We contrast these findings to the study of interactions as determined by the Shapley-Owen value. 2 - Simulated Decision Processes Fabio Maccheroni, Universita Bocconi, Via Sapfath, 25, Milan, Italy, Carlo Baldassi, Simone Cerreia-Vioglio, Massimo Marinacci A successful model to describe two-alternative speeded decisions between consumption goods is the Drift Diffusion Model (DDM) of Ratcliff (1978). In this paper, we provide an axiomatic foundation for the DDM. Our axioms present a test for model misspecification and connect the externally observable properties of choice induced by the DDM with its mechanistic account of the choice process. Moreover, by combining binary DDM comparison with Markovian exploration, we extend the DDM to multi-alternative choice. The resulting “Metropolis-DDM algorithm” is consistent with the eye-tracking findings of Russo and Rosen (1975), converges to the multinomial logit, and is robust to discretization. 3 - Revisiting Ellsberg and Machina’s Paradoxes: A Two Stage Evaluation Model Under Ambiguity Ying He, University of Southern Denmark, Campusvej 55, Odense, M, 5230, Denmark We revisit the Ellsberg paradox by showing that the preference does satisfy the independence axiom under a weak assumption. Such an observation motivates us to develop a two-stage model for decision making under ambiguity. This two- stage model can not only accommodate preferences in different versions of Ellsberg’s paradoxes but also the preferences in paradoxes recent proposed in Machina (2009, 2014) that challenge the validity of many existing models for decision making under ambiguity. Finally, we propose some conjectured examples to illustrate how our model becomes more flexible compared with some existing models in the literature without sacrificing the parsimony of the model.

n MD42 North Bldg 221C Infrastructure Systems Towards a Smart City Emerging Topic: Smart Cities Emerging Topic Session Chair: Hiba Baroud, Vanderbilt University, Nashville, TN, 37235, United States 1 - Analyzing Socially Considerate Multi-Modal Routing Algorithms Chinmaya Samal, Vanderbilt University, Nashville, TN, United States, Abhishek Dubey Ride-sharing platforms such as Uber and Lyft are becoming a more commonplace, particularly in urban environments. While such services may be deemed more convenient than riding public transit due to their on-demand nature, reports show that they do not necessarily decrease the congestion in major cities. In this presentation, we will describe socially considerate multi-modal routing algorithms that are proactive and consider, via predictions, the shared effect of riders on the overall efficacy of mobility services. Our results indicate that even at a low penetration (social ratio), we are able to achieve an improvement in system-level performance. 2 - Modernization of Buildings: The Interplay of Buildings and Smart Cities’ Emerging Systems Atefe Makhmalbaf, University of Texas-Arlington, Arlington, TX, United States Buildings are an integral part of the built environment. Americans spend about 90% of their time indoors. Hence, building design and operation have direct impact on occupant’s health and productivity. Moreover, buildings consume about 40% of energy and 72% of electricity and contribute to 34% of greenhouse gas emissions. Successful deployment of smart cities necessitates design of smart buildings. This presentation will discuss the evolving roles of buildings and the importance of decision support systems during all phases of a building life-cycle from design, to construction and operation to satisfy the broader objectives of smart cities including those of sustainability and resiliency. 3 - Water Infrastructure Resilience: Measurement and Optimization Mohsen Shahandashti, Assistant Professor, University of Texas at Arlington, Arlington, TX, United States A variety of metrics could be used to measure the resilience of water distribution networks subjected to natural hazards, such as earthquakes. Selection of resilience metrics has a significant impact on formulating optimized rehabilitation policies. This significant impact is highlighted by comparing various optimized policies for proactive seismic rehabilitation of a water pipe network based on different resilience metrics. These results highlight the importance of aligning resilience measurement with the experience and expectations of water utilities to formulate effective proactive rehabilitation policies.

n MD43 North Bldg 227B Sandia Session Emerging Topic: Energy and Climate Emerging Topic Session

Chair: Peter Kobos, Sandia National Laboratories, NM, United States Co-Chair: John Eddy, Sandia National Labs, Albuquerque, NM, 87185-1011, United States 1 - The Microgrid Design Toolkit John Eddy, Sandia National Labs, P.O. Box 5800, Albuquerque, NM, 87185-1011, United States The MDT is a visual design and trade-space optimization capability for microgrids. It uses a multi-objective optimization algorithm that executes a discrete event Monte-Carlo simulation to characterize performance and reliability of candidate microgrid designs. The output is a Pareto frontier of efficient alternative microgrid designs and visualizations to help a designer understand the trade-offs. This presentation will describe the technology underlying the MDT and discuss some recent analyses conducted.

n MD41 North Bldg 226C Decision Analysis Society Awards Sponsored: Decision Analysis

Sponsored Session Chair: Jason Merrick 1 - 2018 Decision Analysis Practice Award Competition Saurabh Bansal, Penn State University, 405 Business Building, State College, PA, 16801, United States This session will showcase the finalists for the 2018 Decision Analysis Practice Award Competition. The specific talks in the session will be listed as the finalists are identified and consent to present in the session.

228