2017 IISE Annual Conference and Expo

Simio is excited that the IISE Annual Show will be in our hometown of Pittsburgh, PA. Learn more about all the events we have planned and presentations that used Simio below! Be sure to visit booth #112 to see us.

Saturday, May 20, 2017

Simio Baseball Night

4:05pm at PNC Park
We are hosting a baseball watching party at the Iron City Skull Bar. The Iron City Skull Bar is an outdoor location that offers panoramic views of both the Pirates playing field and the City of Pittsburgh. We are giving away over 30 tickets to IISE attendees! All winners will also receive a Pittsburgh Simio Vintage Tshirt! Learn More!

Sunday, May 21, 2017

Enhancing Primary Care Access with Data Driven Simulation

8:00-8:20am at Room 403
Led by: Zhaohui Geng, Bopaya Bidanda (University of Pittsburgh), Michael Kennedy, Ziyi Kang, Robert Monte (Pittsburgh Veterans Affairs Medical Center)
With the increasing need for responsive primary care access, modeling strategies can help address and improve metrics by testing and optimizing results of various scenarios. While such discrete-event simulation models are typically facility-specific and not designed to be used for various medical centers, we present a ‘blank’ simulator that first modeled the primary care access process. With input from the real-life data and automated analysis, the simulator was used to generate the ‘key’ access metrics of the process driven by the data. In order to make the model generalizable for different medical centers, a procedure will be given to balance the validity and the level of granularities. Finally, main input variables were designed to be controllable by the model users to test ‘what-if’ scenarios.

Simulating Pharmacy Operations to increase Medication Therapy Management

8:20-8:40am at Room 403
Led By: Hamdy Salman and Bryan Norman (University of Pittsburgh)
Patient safety and medication adherence are critical challenges in healthcare environments that have led to numerous interventions to improve safe and effective use of medicines. One effective intervention is providing Medication Therapy Management (MTM) programs to patients. This research focuses on providing more MTM for patients at community pharmacies in order to improve their medication understanding. The key challenge is that MTM can only be provided by a pharmacist who is responsible for oversight of all dispensing of medications in any community pharmacy. Therefore, a simulation model was built to evaluate different prescription filling strategies that affect pharmacist utilization and these strategies are evaluated with the objectives of minimizing patient waiting time, maximizing the number of patients serviced, and providing as much MTM as possible. A number of important variables are taken into account including patient arrival patterns, staff schedules, and prescription abandonment. Prescription abandonment is something that has become a significant issue with the shift towards electronic prescriptions (the number of e-prescriptions received in community retail pharmacies increased from only 29 million in 2007 to over 9.7 billion in 2015). The simulation model is also used to evaluate the value of obtaining varying degrees of prescription information such as

Designing an Extended Stay Unit Using Simulation: A Case Study

8:40-8:20am at Room 403
Led By: Mohammad Al-Mashraie (Binghamton University)
Healthcare systems are facing an increasing demand as more and more patients are becoming insured. This increase ]has made it vital for these systems to be managed effectively and efficiently. One of the characteristics of healthcare facilities, such as hospitals, is the limited resources that has to be managed in a way that enables providing patients with higher value in a lower cost. In this paper, a case study is presented in which Discrete Event Simulation (DES) is utilized to design an extended stay unit for outpatients after their surgery. The uncertainty in patients’ length of stay, arrival rate, and the characteristics of the locations that these patients could arrive from was modeled. After defining the patients who are qualified to go to this unit, a year worth of data was collected from an academic medical center to be used to make a data driven decision on the number of beds required based on certain performance measures such as patients waiting times and utilization. The results of this study will help in avoiding operational inefficiencies before the unit comes alive as well as testing several what-if scenarios that are related to the design of this unit.

Probabilistic methods for long-term demand forecasting for aviation production planning

12:30-12:50pm at Room 318
Led By: Minxiang Zhang, Cameron MacKenzie, Caroline Krejci, John Jackman, Guiping Hu, (Iowa State University), Charles Hu, Adam Graunke, and Gabriel Burnett (Boeing Company)
Long-term demand forecasting is critical to the aviation industry since firms need to make important decisions with intensive capital investment. Accurately forecasting demand is challenging because of the significant variability and the large number of plausible futures. The application of probabilistic methods in forecasting time series has been well studied especially in the financial industry. This presentation will discuss two models: (1) a novel application of Brownian motion in order to account for dependency between observations and (2) a forecast using geometric Brownian motion. The results from these models will be compared to an autoregressive integrated moving average model. The modified Brownian motion and the geometric Brownian motion will be applied to forecasting demand for aircraft 20 years into the future.

Capacity Planning and Production Scheduling for Aircraft Painting Operations

12:50-1:10pm at Room 318
Led By: Xiangzhen Li, Caroline Krejci, Cameron MacKenzie, John Jackman, Guiping Hu (Iowa State University), Charles Hu, Adam Graunke, and Gabriel Burnett (Boeing Company)
Long-term capacity planning and production scheduling present significant challenges for the aviation industry. Our research has integrated three different modeling methodologies in an effort to effectively forecast future demand for aircraft painting and then assess and manage the capacity that is needed to meet these requirements. First, an innovative forecasting approach was developed in which the standard Brownian motion parameter estimation technique was modified to allow for dependencies between demand observations in sequential time periods. These demand forecasts become inputs to an integer programming model, which is used to generate optimal monthly aircraft painting schedules. This allows for capacity assessment that is based on optimal scheduling, rather than the existing heuristic-based methods. The optimal monthly schedules then serve as inputs into a discrete event simulation model of the painting operation, which is used to test the robustness of the optimal schedules to uncertain demand and processing times. The outputs of this simulation model provide a basis for estimating system risk, including the likelihood of delivery tardiness and exceeding existing capacity.

Simulation of Distribution Centers using Simio - Case Study

3:50-4:10pm at Room 405
Led By: Dusan Sormaz, Alejandro Romero Montoya, Brian Williams, Joseph Weiser, and Esteban Rodriguez (Ohio University)
In this paper we present a simulation study of a high volume distribution center . The model is developed for the delivery side of distribution. The flow starts with receiving orders and continues to the picking areas, where the items of the order are picked from one of three multi-level, multi-stage storage areas. All storage areas are connected by high-speed conveyors and the order follows the conveyors through all storage areas. Each stage in each storage ara has dedicated item pickers, and one of goals was to investigate their assignments and schedules. The features of Simio are used to develop and implement the high fidelity animated model. The DC performance is measured in cycle time of orders, throughput and picker utilizations. Few scenarios are investigated and the corresponding recommendations are given.

Simulation of Car Light Assembly Cell - Case Study

4:10-4:30pm at Room 405
Led By: Dusan Sormaz, James Lefebvre, and Melinda Nelson (Ohio University)
In this paper we present a simulation study of a car light assembly cells. The model is developed for the self-contained assembly cell which contains all resources and several operators. The assumption in the model was that all parts for the light assembly are available in sufficient inventories. The model addressed the work distribution between operators. The assembly cell has 14 work stations which are operated by 4-6 operators. The features of Simio are used to develop and implement the high fidelity animated model. The cell performance is measured in cycle time of car light orders, throughput and operator utilizations. The relationships between number of operators and their work assignments, on one side and the system performance was studied by simulation experiments. Few scenarios were investigated and the corresponding recommendations were given.

Application of Iterative Optimization-based Simulation (IOS) approach in Supply Chain Management (SCM) Systems

5:00-6:40pm at Room 415
Led By: Mohammad Dehghanimohammadabadi (Northeastern University) and Thomas K. Keyser (Western New England University)
Supply chain models are usually stochastic with non-linear and complex relationships. In these models, customers’ demand and suppliers’ capacities are constantly changing both in terms of variety and price range. Moreover, supply chain relationships between suppliers, manufacturers, distributors and retailers evolve over time. Therefore, a simulation model is introduced which embraces diverse SCM models. This model considers many of the above-mentioned changes in the supply chain as a trigger point to reconfigure the system. Applying the proposed Iterative Optimization-based Simulation (IOS) approach provides companies an opportunity to design their own supply chain system, not only by optimizing internal operations but also by examining and improving the entire supply chain performance over the long-term.

University Parking System Analysis through Discrete Event Simulation

5:20-5:40pm at Room 405
Led By: Alaa Alghwiri, Shengyong Wang, and Jared Coleman (University of Akron)
This research initiative focuses on modeling and optimizing an urban university parking system. Different from regular parking systems, urban university parking systems face additional challenges due to the limited space, the complexity of the parking facility distribution, the dynamic nature of the traffic flows, and the randomness of the parking access times, primarily driven by class schedules. A comprehensive discrete event simulation model was developed to capture the parking system dynamics for an urban university in Ohio. Specifically, a model consisting of 7 parking decks, 44 surface lots, and 8 residence halls was created to take into account the different user groups, their arrival and departure patterns, and parking location preferences. After validating the baseline model, an interactive campus map was developed to show the best places to park throughout the day based on the windows of arrival times on campus. This map can be updated each academic semester based on the arrival rate data for parking location recommendations.

Modeling and Optimizing University Shuttle Bus System through Bus Diversions

5:40-6:00pm at Room 405
Led By: Alaa Alghwiri, Shengyong Wang, and Jared Coleman (University of Akron)
Different than regular public transportation systems, the transportation system in an urban university setting poses additional challenges due to the dynamic nature of the user groups (mainly students), as many factors including the living arrangements and class schedules of the students influence the demands for on-campus transportation. In this research initiative, a university shuttle system was simulated using discrete event simulation modeling to capture the systems dynamics, identify system bottlenecks, and recommend improvement strategies. Specifically, as the key performance measures, the passenger waiting times and the shuttle bus utilization were analyzed at each bus stop. After modeling and validating the baseline model, a hypothetical system improvement scenario of diverting buses from one route to another was investigated. Subsequently, the optimal bus diversion policy in terms of diversion trigger points, timing, and the number of diverted buses was recommended.

Monday, May 22, 2017

Comparison of Alternative Hospital Supply Chain Systems using Simulation

8:00-8:20am at Room 306
Led By: Michael Kuhl and Siddharth Garg (Rochester Institute of Technology)
Hospital supply chain systems play an important role in the delivery of high quality patient care. Having the right products available at the point-of-use is important to the efficient and effective treatment of patients. Although there is a vast quantity of current supply chain research, much of the literature focuses on particular aspects of the supply chain. In this research, we study the hospital supply chain from manufacturers/distribution centers to the point-of-use within a hospital unit, taking into account the integration and implementation of the various echelons of the supply chain system. In particular, we design and compare alternative supply chain systems including a par level and Kanban systems. We utilize a simulation and optimization methodology to evaluate supply chain decision variables (order quantity, safety stock, etc.) and to compare alternative system configurations based on service level and operational costs, subject to variability in demand and lead-time, as well as perishable product and inventory space constraints.

Hospital Patient Flow Capacity Planning Simulation Model at Vancouver Coastal Health

8:20-8:40am at Room 309
Led By: Ernest Wu and Amanda Yuen (Vancouver Coastal Health)
In order to address certain operational questions for a hospital, Vancouver Coastal Health built a simulation model that captures the flow of all patients from multiple arrival streams all the way to discharge. For this model, historical data was analyzed to determine arrival rates, transfer rates between units, and processing times within each area of the hospital. Different patient types were created to represent the different needs and journeys of patients through the hospital. The actual inpatient bed capacities and operating room schedules were used, and additional process logic was created to simulate certain system dynamics such as surgical cancellations and the use of overflow capacity. This model helped address questions such as the impact of implementing protected surgical beds and discharging additional patients from certain units.

Coexistance of a Simulation Model and Real-Time Planning & Scheduling

12:30-1:50pm at Room 301
Led By: Renee Thiesing (Simio LLC)
Come learn how a simulation model that can be used for facility design or business process optimization can also be used for real-time planning and scheduling. We will discuss how one Simio tool is a dual-purpose solution that can be used to address future planning decisions as well as daily, operational decisions. We will explore how Simio now leverages the cloud computing power of Microsoft Azure to support your most demanding applications through the Simio Portal Edition. Simio Portal Edition can be used for rapid experimentation or to simply share results across your enterprise. Let us show you an overview of the NEW Simio experience and see why we are always “Forward Thinking.”

Simulation of the Implementation of Lean Tools in Laboratory Post Press

3:50-4:10pm at Room 319
Led By: Luis Daniel Palacios Vázquez, Carlos Quintero Aviles, and Isrrael González Nuñez (Universidad Tecnologica Fidel Velaquez)
In this paper the simulation of the benefits that can be obtained through the implementation of lean manufacturing tools in a didactic production process performed in the laboratory of post press Technological University Fidel Velázquez (UTFV) is performed. The process involves the development of "Booklets" where students play the role of representing a company that makes these booklets and must comply with the orders requested by the customer; in a first stage is process is done by creating chaos physically between areas and positions that do not achieve their objective at this stage operations data, distances and routes with which the simulation model is created are taken, this model serves students to propose and analyze changes to improve the process before physically performing the following steps.

The difference of mean waiting times between two classes of customers in a single-server FIFO queue: an experimental study

5:00-5:20pm at Room 303
Led By: Rodrigo Romero-Silva and Margarita Hurtado (Universidad Panamericana)
Previous studies have shown that the distribution of the mean waiting time of different classes of customers is different for each class of customer in a GI/G/1 queue with FIFO discipline. This experimental study is motivated by those results as it investigates in which conditions does a difference of mean waiting times between two classes of customers are present using a FIFO discipline. Results from the study show that the most important factor to determine whether a difference of mean waiting times between two classes exists or not is the presence of a difference of squared coefficients of variation of inter-arrival times between both classes. Furthermore, if a difference between mean service times of both classes exists, the class of customers with the highest mean service time will tend to have the smallest mean waiting time of both classes.

Simulation of a Disruption in a Multi-Vendor Supply Chain

5:20-5:40pm at Room 305
Led By: Pranav Joshi (SUNY Binghamton)
A minute change at the vendor level can be responsible for the disruption of an entire supply chain like a stock-out situation. For a smooth operation of the supply chain, conventional approach has been to support the system with multiple vendors or backups, which sacrifice the overall operational cost. Therefore, it is important to determine proper number of vendors to keep the supply chain resilient. Previous study like robust optimization find an optimal solution while incorporating some uncertainties. However, the complexity of the model is not a trivial issue if we want to incorporate operations within a vendor at process level. Detailed processes inside vendors will add more uncertainty into the model which reflects an entire supply chain system. Therefore, a simulation model considering multiple vendors has been studied to investigate the effect of a disruption in a multiple-vendor supply chain. Moreover, the benefit of this simulation study lies in making decision of how to deal with the situation when a disruption occur. Some numerical examples of the supply chain disruption simulation at the process level will be provided to compare different multi-vendor policies.

The difference of mean waiting times between two classes of customers in a single-server FIFO queue: an experimental study

5:40-6:00pm at Room 303
Led By: Rodrigo Romero-Silva, Margarita Hurtado (Universidad Panamericana), Sabry Shaaban (ESC La Rochelle), and Erika Marsillac (Old Dominion University)
Previous studies have shown that the mean queue length of a GI/G/1 system is significantly influenced by the skewness of inter-arrival times but not by the skewness of service times. These results were limited by the fact that all considered distributions were positively skewed and had high values of squared coefficient of variation (more than 1). The purpose of this paper is to investigate which is the effect that the skewness of both inter-arrival and service times has on the probability distribution of waiting times when a negatively skewed distribution with low values of squared coefficient of variation (less than 1) is considered to model inter-arrival and service times. After conducting a series of experiments of a GI/G/1 queue using discrete simulation results showed that the lowest mean waiting time and the lowest coefficient of variation for the waiting times can be attained with a combination of positive inter-arrival skewness and negative service skewness when the coefficient of variation of inter-arrival times is lower than 1.

Tuesday, May 23, 2017

Model simulation evacuation of people from buildings for tsunami disaster

8:20-8:40am at Room 318
Led By: Raul Zuñiga, Pablo Opazo, Alfredo Fuentes and Matias Contreras (Universidad Arturo Prat)
Chile is one of the most seismically active countries in the world, where natural disasters occur more frequently, such as earthquakes and tsunamis. These affect to people who stay in buildings closer to the sea. For this reason is necessary to have effective evacuation plans in order to save lives. The aim of this paper is to develop a simulation model to calculate evacuation delay time of people from the buildings located in susceptible areas. The methodology considers a case study of four buildings at Arturo Prat University, located in Iquique city, Chile. The model is developed by using SIMIO simulation software. This model is the basis for future research because there are many buildings very close to the sea and there are people with differents kind of physical limitations.The model will allows to improve the evacuation plans of the Iquique city.

Patient Scheduling at Breast Center Using Discrete Event Simulation

11:20-11:40am at Room 402
Led By: Mandana Rezaeiahari, Linda Adwan, and Mohammad Khasawneh (Rochester Institute of Technology)
This research addresses the scheduling of breast center patients at a local hospital with the purpose of minimizing patient flow time while maintaining patient throughput. The initial scheduling policy was based on First-Come-First-Serve (FCFS) rule, which resulted in staff overtime and long patient flow time. Therefore, a patient mix strategy is used to address this problem. Also, to show the effectiveness of the patient case mix strategy, shortest processing time (SPT) scheduling rule is used to compare patient throughput and average flow time in the system. In this study, there are two types of patients: follow-up and consult patients. Discrete event simulation is used to model the scheduling problem over a four-month period. Based on the experimental analysis and results, case mix resulted in the best outcomes with 85% and 65% average reduction in waiting time for follow-up and consult patients compared to the baseline, respectively. On the other hand, patients stay significantly longer in the system under the SPT rule (follow-ups: 10.5%; new consults: 48.3%) compared to the case mix scenario.

Simulating Airport Terminals: Implementations around the Globe

12:30-12:50pm at Room 308
Led By: Aarshabh Misra and Daley Mikalson (ARUP)
Airport terminal buildings are complex facilities that consist of numerous components, each critical to ensure its safe and efficient operation. Inevitably, complex planning approaches are required to inform capital expenditure, when considering infrastructure upgrades and phasing.
Discrete event simulation can be employed to model complex passenger behavior and movement in the airport environment to test facility performance, generate infrastructure requirements for passenger processing, and identify constraints in existing facilities as well as expansion programs. Iterative development of simulations allows the design to be progressively refined and informs evidence based decision making.
At Arup, simulation and modelling has been employed on major airport projects around the world including in Beijing, Abu Dhabi, New York JFK, Portland, Seattle and Manchester, amongst others. Various passenger processes such as conventional and kiosk check-in, bag induction, security clearance, immigration, customs and baggage hall have been simulated. Case studies are presented to discuss the general modelling concept, primary inputs, methodology of the analyses performed, and the outcomes for process flow improvement in airports. Implications of simulation and modelling are significant, and are described in the context of airport planning and policy.

Utilizing Simulation Modeling to Test Greenfield Design of Healthcare Product Production Equipment

2:00-2:20pm at Room 319
Led By: Amy Greer (Mosimtec)
A healthcare company developed a novel process for creating a prescription based product utilized by a large number of consumers. The company then worked with mechanical and electrical engineers to design equipment that would be capable of automatically performing this process with a throughput capability high enough to justify investing in this approach. A simulation model was developed to test initial designs.
The simulation modeling engagement uncovered a previously undetected airspace conflict between two moving parts in the equipment. This airspace conflict, and previously planned control logic, would mean desired throughput capabilities would not be met with the initial design. The design team made physical changes to the equipment, along with exploring system timing improvements that could be made with the control logic. These changes were implemented and analyzed in the simulation model, with the simulation team able to provide recommendations and insights on how to further improve system performance.
Attendees will see how traditional Industrial Engineering simulation modeling can be extremely valuable in designing medical devices, even in systems with little or no variability. The approach for modeling airspace conflicts and the discrete event – agent based architecture will be presented.

A Decision-Making Model for Optimizing Schedules of Aerospace Subsystems

2:00-2:20pm at Room 308
Led By: Samar Alhihi, Daniel Trembley (The State University of New York at Binghamton), and Sarah Lam (Binghamton University)
The aerospace industry requires on-time deliveries of high quality and reliable avionic subsystems. The difficulties of manufacturing in a low-rate built-to-order environment of avionic subsystems often derive from outdated methodologies of scheduling. The manufacturing and production activities that coincide with a contractual delivery date are often ineffective. Since many aerospace manufacturers build the subsystem kits that share the same flowlines, this leads to disruption through competition of manufacturing resources and contributes to missed deliveries. The uneven manufacturing and production flowlines (through partial builds, vendor delivery schedules, machining constraints, and labor availability) become contributors to a propagated delay in critical paths of the master schedule. The decision-making process for setting build priorities, in many cases, is not well defined, and often relies on a scheduler’s experience. This research models the decision-making process in such uncertain and complex environments. The model identifies and predicts schedule risk at all levels of the manufacturing and production process. Using the Simio optimization software tool OptQuest with common job shop Enterprise Resource Planning (ERP) data to assess this risk, the research method combines the predicted scheduled delivery dates, and formulates through sensitivity analysis, items and activities within the schedule that cause concern.

Agent-based Modeling and Simulation

4:50-5:10pm at Room 319
Led By: Thomas Kehl and Andreas Rinkel (HSR)
Agent-based modeling and simulation is a quite modern approach to model systems. It allows modeling of the dynamics of complex and cybernetic systems. These are often self-organizing systems which produce emergency effects, e.g. the escape behavior of people. In this paper, we first are going to give an overview of agent-based modeling methods and define based on our survey a set of concepts and capabilities to constitute a basic set of model components to establish a library for an agent-based modeling approach.
In the second step, we show how to implement such a library in Simio; an object oriented professional simulation framework. The proof of concept is shown in a simple pedestrian model that simulates the escape behavior of pedestrians, caused by heat sources.

Wednesday, May 24 - Thursday May 25, 2017

Simio User Group Meeting

Pittsburgh Drury Hotel
Simio is excited to announce that it will hold its first user group meeting in Pittsburgh, PA on May 24-25th, 2017. This meeting will follow the 2017 Institute of Industrial and Systems Engineers Annual Conference and Expo that is also in Pittsburgh. We invite you to join other users and friends of Simio in a various industries and companies such as: Boeing, Lockheed Martin Aeronautics, HDR Architecture, Sun Chemical S.A., Honeywell FMT, American Airlines, Array Advisors, and many more! Learn More!