Detailed Discrete Event Simulation Model Research Articles

Maximizing hospital capacity to serve pandemic patient surge in hot spots via queueing theory and microsimulation

This paper presents a mixed-integer mathematical program with embedded equations developed from concepts of queueing theory and Jackson networks for estimating the steady-state maximum potential hospital capacity for COVID-19 patient care in extreme surge conditions, where a hospital must turn nearly all of its existing resources toward the care of pandemic patients. Estimating the potential maximum hospital capacity for pandemic patient care can aid in assessing regional healthcare capacity during surges in pandemic patient demand, predicting shortfalls, and designing preparedness and response actions. To obtain such estimates and inform action, the program determines a best assignment of a heterogeneous staff of nurses and doctors to key units appropriate for their skills to create the optimal allocation of staffed beds. An alternative trial-and-error approach is offered that decision-makers without optimization or software expertise can use to obtain similar estimates. Under a given assignment of resources, a variety of key performance indicators can be obtained through the direct use of the queueing equations. Results of comparisons to outcomes from a detailed discrete event simulation model of an identical hospital design show the accuracy of the equations to be high despite the added simplifications needed for the use of a closed-form equation-based methodology.

Trade-offs in the landside operations of air cargo hubs: Horizontal cooperation and shipment consolidation policies considering capacitated nodes

Landside operations in air cargo terminals consist of many freight forwarders (FFWs) delivering and picking up cargo at the capacity-constrained loading docks at the airport's ground handlers' (GHs) facilities. To improve the operations of the terminal and take advantage of their geographical proximity a small set of FFWs can build a coalition to consolidate stochastically-arriving shipments and share truck fleet capacity while other FFWs continue bringing cargo to the terminal in a non-cooperative manner. Results from a detailed discrete-event simulation model of the cargo landside operations in Amsterdam Aiport showed that all operational policies had trade-offs in terms of the average shipment cycle time of coalition FFWs, the average shipment cycle time of non-coalition FFWs, and the total distance traveled by the coalition fleet, suggesting that horizontal cooperation in this context was not always beneficial, contrary to what previous studies on horizontal cooperation have found. Since dock capacity constitutes a significant constraint on operations in air cargo hubs, this paper also investigates the effect of dock capacity utilization and horizontal cooperation on the performance of consolidation policies implemented by the coalition. Thus, we built a general model of the air cargo terminal to analyze the effects caused by dock capacity utilization without the added complexity of landside operations at Amsterdam Airport to investigate whether the results hold for more general scenarios. Results from the general simulation model suggest that, in scenarios where dock and truck capacity become serious constraints, the average shipment cycle times of non-coalition FFWs are reduced at the expense of an increase in the cycle times of FFWs who constitute the coalition. A good balance among all the performance measures considered in this study is reached by following a policy that takes advantage of consolidating shipments based on individual visits to GH.

Digital Twin for Training Bayesian Networks for Fault Diagnostics of Manufacturing Systems.

Smart manufacturing systems are being advocated to leverage technological advances that enable them to be more resilient to faults through rapid diagnosis for performance assurance. In this paper, we propose a co-simulation approach for engineering digital twins (DTs) that are used to train Bayesian Networks (BNs) for fault diagnostics at equipment and factory levels. Specifically, the co-simulation model is engineered by using cyber–physical system (CPS) consisting of networked sensors, high-fidelity simulation model of each equipment, and a detailed discrete-event simulation (DES) model of the factory. The proposed DT approach enables injection of faults in the virtual system, thereby alleviating the need for expensive factory-floor experimentation. It should be emphasized that this approach of injecting faults eliminates the need for obtaining balanced data that include faulty and normal factory operations. We propose a Structural Intervention Algorithm (SIA) in this paper to first detect all possible directed edges and then distinguish between a parent and an ancestor node of the BN. We engineered a DT research test-bed in our laboratory consisting of four industrial robots configured into an assembly cell where each robot has an industrial Internet-of-Things sensor that can monitor vibrations in two-axes. A detailed equipment-level simulator of these robots was integrated with a detailed DES model of the robotic assembly cell. The resulting DT was used to carry out interventions to learn a BN model structure for fault diagnostics. Laboratory experiments validated the efficacy of the proposed approach by accurately learning the BN structure, and in the experiments, the accuracy obtained by the proposed approach (measured using Structural Hamming Distance) was found to be significantly better than traditional methods. Furthermore, the BN structure learned was found to be robust to variations in parameters, such as mean time to failure (MTTF).

A two-level optimisation-simulation method for production planning and scheduling: the industrial case of a human–robot collaborative assembly line

In this work, a novel optimisation-simulation based on the Recursive Optimisation-Simulation Approach (ROSA) methodology is developed to provide effective decision-support for integrated production planning and scheduling. The proposed iterative approach optimises production plans while satisfying complex scheduling constraints, such as robots' allocation in collaborative tasks. The plans are determined through a two-level MILP model and are iteratively evaluated by a detailed discrete-event simulation model to guarantee capacity-feasible solutions at the scheduling level. Through an industrial case study of a multistage assembly line design collaboratively operated by humans and mobile shared robots, near-optimal solutions comprise lot-sizing decisions, the release schedule of production orders, the allocation of tasks to humans or robots, and the number of robots per period. Moreover, by addressing a set of propositions to assess the methodology, the results highlight the advantages of the hybrid approach to converge into optimised operational decisions and analyse the process dynamics.

Addressing artificial variability in patient flow

Addressing artificial variability in patient flow is an effective approach to improving accessibility and quality of care and reducing waste and cost in healthcare systems. The most significant artificial variability in patient flow is attributable to dysfunctionally scheduled admissions that can be decreased or eliminated by load smoothing. In this study, we examine the impact of load smoothing of scheduled admissions on patient flow performance metrics of an obstetric unit to provide insights for capacity management. We also investigate the relationship between the impact of load smoothing of scheduled admissions on the patient flow performance metrics of the unit and the volume of unscheduled admissions. In doing so, we develop a detailed discrete-event simulation model of the patient flow of the unit in which patients are categorized into different classes. We compare the results of the simulation model before and after implementing different degrees of load smoothing by considering various ratios of average weekend daily load to average weekday daily load. We determine how different degrees of load smoothing reduce the number of beds required, the average waiting time, and the average probability of delay differently while they have no considerable impact on the average bed occupancy rate. Moreover, considering different volumes of unscheduled admissions, we quantify how the reduction in the average waiting time and the average probability of delay by load smoothing increases when the ratio of unscheduled admissions to scheduled admissions decreases.

The impact of implementing a demand forecasting system into a low-income country's supply chain

ObjectiveTo evaluate the potential impact and value of applications (e.g. adjusting ordering levels, storage capacity, transportation capacity, distribution frequency) of data from demand forecasting systems implemented in a lower-income country's vaccine supply chain with different levels of population change to urban areas. Materials and MethodsUsing our software, HERMES, we generated a detailed discrete event simulation model of Niger's entire vaccine supply chain, including every refrigerator, freezer, transport, personnel, vaccine, cost, and location. We represented the introduction of a demand forecasting system to adjust vaccine ordering that could be implemented with increasing delivery frequencies and/or additions of cold chain equipment (storage and/or transportation) across the supply chain during varying degrees of population movement. ResultsImplementing demand forecasting system with increased storage and transport frequency increased the number of successfully administered vaccine doses and lowered the logistics cost per dose up to 34%. Implementing demand forecasting system without storage/transport increases actually decreased vaccine availability in certain circumstances. DiscussionThe potential maximum gains of a demand forecasting system may only be realized if the system is implemented to both augment the supply chain cold storage and transportation. Implementation may have some impact but, in certain circumstances, may hurt delivery. Therefore, implementation of demand forecasting systems with additional storage and transport may be the better approach. Significant decreases in the logistics cost per dose with more administered vaccines support investment in these forecasting systems. ConclusionDemand forecasting systems have the potential to greatly improve vaccine demand fulfilment, and decrease logistics cost/dose when implemented with storage and transportation increases. Simulation modeling can demonstrate the potential health and economic benefits of supply chain improvements.

Removing the regional level from the Niger vaccine supply chain

ObjectiveSince many of the world's vaccine supply chains contain multiple levels, the question remains of whether removing a level could bring efficiencies. MethodsWe utilized HERMES to generate a detailed discrete-event simulation model of Niger's vaccine supply chain and compared the current four-tier (central, regional, district, and integrated health center levels) with a modified three-tier structure (removing the regional level). Different scenarios explored various accompanying shipping policies and frequencies. FindingsRemoving the regional level and implementing a collection-based shipping policy from the district stores increases vaccine availability from a mean of 70–100% when districts could collect vaccines at least weekly. Alternatively, implementing a delivery-based shipping policy from the central store monthly in three-route and eight-route scenarios only increases vaccine availability to 87%. Restricting central-to district vaccine shipments to a quarterly schedule for three-route and eight-route scenarios reduces vaccine availability to 49%. The collection-based shipping policy from district stores reduces supply chain logistics cost per dose administered from US$0.14 at baseline to US$0.13 after removing the regional level. ConclusionRemoving the regional level from Niger's vaccine supply chain can substantially improve vaccine availability as long as certain concomitant adjustments to shipping policies and frequencies are implemented.

Impact of Recipe Restrictions on Photolithography Toolsets in an ASIC Fabrication Environment

In this paper, a detailed discrete event simulation model is used to better understand the effects of recipe constraints resulting from process restrictions and tool capabilities on overall average cycle time performance of a photolithography area and on average cycle times of individual mask layers. The study is motivated by the industry, in which engineers have to frequently make decisions on tool qualifications and recipe coverage. An experimental procedure is developed and implemented to show the impact of different levels of tool paths on photolithography toolsets. The simulation results show that increasing the number of tool paths decreases the overall average photolithography cycle time for particular wafer loading levels. Also, as start volumes increase, toolset utilizations increase and the impact of single-path tools on average cycle times increases. Immature processes and low-use processes tend to have more single paths and thus suffer higher average cycle times accordingly. Furthermore, it is reported that average cycle time decreases significantly under multiple process environments due to the lower impact of single paths.

Reliable Buffering Management Algorithm Support for Multicast Protocol in Mobile Ad-hoc Networks

Multicasting is one of the relevant issues of communication in infrastructure or centralized administration networks. The reliable delivery of multicast data packets needs feedback from all multicast receivers to indicate whether a retransmission is necessary. A reliable multicast delivery in the wireless Ad-hoc network requires a multicast packet to be received by all multicast receiver nodes. Thus, one or all members need to buffer data packet for possible error recovery. Furthermore, different buffer strategies are essentially used in existing reliable multicast protocols towards support error recovery and reducing buffer overflow. This study proposed two algorithms to improve the performance of the source tree reliable multicast (STRM) protocol. The first algorithm was developed to avoid buffer overflow in the sender node as the forward server (FS) nodes of STRM. This reduction is achieved by managing the buffer of the FS nodes, i.e., selecting the FS nodes depending on its empty buffer size and reducing the feedback sent from the receiver nodes to their FS node. The second algorithm was developed to decrease duplicated packets in the multicast members of the local group, which may be achieved by sending the repair packets to the requesting member. The FS in the local group should create a dynamic and temporary subgroup whose members are only those that requested the repair packet retransmission. The algorithms were tested using detailed discrete event simulation models encompassing messaging systems including error, delay, and mobility models to characterize the performance benefits of the proposed algorithms compared with the existing wireless Ad-hoc network protocols. Several experiments were conducted, revealing numerous results that verify the superior performance of the proposed algorithms over the existing algorithms.

Coping with disassembly yield uncertainty in remanufacturing using sensor embedded products

This paper proposes and investigates the use of embedding sensors in products when designing and manufacturing them to improve the efficiency during their end-of-life (EOL) processing. First, separate design of experiments studies based on orthogonal arrays are carried out for conventional products (CPs) and sensor embedded products (SEPs). In order to calculate the response values for each experiment, detailed discrete event simulation models of both cases are developed considering the precedence relationships among the components together with the routing of different appliance types through the disassembly line. Then, pair-wise t-tests are conducted to compare the two cases based on different performance measures. The results showed that sensor embedded products improve revenue and profit while achieving significant reductions in backorder, disassembly, disposal, holding, testing and transportation costs. While the paper addresses the EOL processing of dish washers and dryers, the approach provided could be extended to any other industrial product.

Impact of changing the measles vaccine vial size on Niger's vaccine supply chain: a computational model

BackgroundMany countries, such as Niger, are considering changing their vaccine vial size presentation and may want to evaluate the subsequent impact on their supply chains, the series of steps required to get vaccines from their manufacturers to patients. The measles vaccine is particularly important in Niger, a country prone to measles outbreaks.MethodsWe developed a detailed discrete event simulation model of the vaccine supply chain representing every vaccine, storage location, refrigerator, freezer, and transport device (e.g., cold trucks, 4 × 4 trucks, and vaccine carriers) in the Niger Expanded Programme on Immunization (EPI). Experiments simulated the impact of replacing the 10-dose measles vial size with 5-dose, 2-dose and 1-dose vial sizes.ResultsSwitching from the 10-dose to the 5-dose, 2-dose and 1-dose vial sizes decreased the average availability of EPI vaccines for arriving patients from 83% to 82%, 81% and 78%, respectively for a 100% target population size. The switches also changed transport vehicle's utilization from a mean of 58% (range: 4-164%) to means of 59% (range: 4-164%), 62% (range: 4-175%), and 67% (range: 5-192%), respectively, between the regional and district stores, and from a mean of 160% (range: 83-300%) to means of 161% (range: 82-322%), 175% (range: 78-344%), and 198% (range: 88-402%), respectively, between the district to integrated health centres (IHC). The switch also changed district level storage utilization from a mean of 65% to means of 64%, 66% and 68% (range for all scenarios: 3-100%). Finally, accounting for vaccine administration, wastage, and disposal, replacing the 10-dose vial with the 5 or 1-dose vials would increase the cost per immunized patient from $0.47US to $0.71US and $1.26US, respectively.ConclusionsThe switch from the 10-dose measles vaccines to smaller vial sizes could overwhelm the capacities of many storage facilities and transport vehicles as well as increase the cost per vaccinated child.

Evaluating the impact of sensor-embedded products on the performance of an air conditioner disassembly line

Increasing consumer awareness towards environmental issues and stricter environmental legislation have forced many manufacturers to set up facilities for product recovery which involves the minimization of the amount of waste sent to landfills by recovering materials and components from returned or end-of-life (EOL) products. Disassembly is an important process in product recovery since it allows for the selective separation of desired parts and materials. EOL products involving missing and/or nonfunctional components increase the uncertainty associated with disassembly yield. Testing, a common solution method, results in high costs. Moreover, if the component is found to be defective, the disassembly time is wasted. Sensor-embedded products (SEPs) can deal with this uncertainty by providing information on the condition of components prior to disassembly. This study evaluates the impact of SEPs on the various performance measures of an air conditioner (AC) disassembly line controlled by a multikanban system which effectively manages material flows considering the stochastic behavior of the disassembly line. First, separate design-of-experiments studies based on orthogonal arrays are carried out for conventional products (CPs) and SEPs. In order to calculate the response values for each experiment, detailed discrete-event simulation models of both cases are developed, considering the precedence relationships among the components of an AC. Then, pairwise t tests are conducted to compare two cases based on different performance measures. The test results show that SEPs improve revenue and profit while achieving significant reductions in backorder, disassembly, disposal, holding, testing, and transportation costs.

Comparison of economic benefits of sensor embedded products and conventional products in a multi-product disassembly line

Sensors embedded into products during the production process are important data acquisition tools for after-sale products’ condition monitoring. By facilitating data collection from critical components in a product, these sensors help companies predict components and/or products failures during product usage. They are also very useful at the end-of-life (EOL) of products. Using sensor information, the conditions, types and remaining lives of components in an EOL product can be determined prior to actual disassembly. In this study, we assess the use of sensors in determining the steps in EOL processing of products. In particular, we evaluate the impact of sensor embedded products (SEPs) on various performance measures of an appliance disassembly line controlled by a multi-kanban system. First, separate design of experiments studies based on orthogonal arrays are carried out for conventional products (CPs) and SEPs. In order to calculate the response values for each experiment, detailed discrete event simulation models of both cases are developed considering the precedence relationships among the components together with the routing of different appliance types through the disassembly line. Then, pair-wise t-tests are conducted to compare the two cases based on different performance measures. The test results show the superiority of SEPs over CPs with respect to all costs (viz., disassembly, disposal, testing, backorder, transportation, holding) as well as revenue and profit.

A prediction interval-based approach to determine optimal structures of neural network metamodels

Neural networks have been widely used in literature for metamodeling of complex systems and often outperform their traditional counterparts such as regression-based techniques. Despite proliferation of their applications, determination of their optimal structure is still a challenge, especially if they are developed for prediction and forecasting purposes. Researchers often seek a tradeoff between estimation accuracy and structure complexity of neural networks in a trial and error basis. On the other hand, the neural network point prediction performance considerably drops as the level of complexity and amount of uncertainty increase in systems that data originates from. Motivated by these trends and drawbacks, this research aims at adopting a technique for constructing prediction intervals for point predictions of neural network metamodels. Space search for neural network structures will be defined and confined based on particular features of prediction intervals. Instead of using traditional selection criteria such as mean square error or mean absolute percentage error, prediction interval coverage probability and normalized mean prediction interval will be used for selecting the optimal network structure. The proposed method will be then applied for metamodeling of a highly detailed discrete event simulation model which is essentially a validated virtual representation of a large real world baggage handling system. Through a more than 77% reduction in number of potential candidates, optimal structure for neural networks is found in a manageable time. According to the demonstrated results, constructed prediction intervals using optimal neural network metamodel have a satisfactory coverage probability of targets with a low mean of length.

Optimizing scheduling of post‐earthquake electric power restoration tasks

AbstractThis paper presents a stochastic integer program developed to determine how to schedule inspection, damage assessment, and repair tasks so as to optimize the post‐earthquake restoration of the electric power system. The objective of the optimization is to minimize the average time each customer is without power, and a genetic algorithm is used to solve it. The effectiveness of the schedules recommended by the optimization are evaluated by running a detailed discrete event simulation model of the restoration process with both the optimization‐generated schedules and the power company's original schedules, and comparing the resulting restorations according to three measures—average time each customer is without power, time required to restore 90 of customers, and time required to restore 98 of customers. The optimization and simulation models both consider all the earthquakes that could affect the power system and represent the uncertainty surrounding expected restoration times. he models were developed through an application to the Los Angeles Department of Water and Power (LADWP) electric power system, but the general approach is extendable to other electric power systems, other lifelines, and other hazards. Copyright © 2006 John Wiley & Sons, Ltd.

Analyzing throughput and capacity of multiproduct batch processes

A detailed discrete-event simulation model is extensively utilized by a major chemical process facility to evaluate new plant designs. This model is further exploited using known statistical techniques to provide valuable sensitivity analysis feedback to engineers for process analysis and improvement. A framework is proposed to guide the implementation and application of multivariate statistics and simulation metamodels to analyze simulation output of multiproduct batch processes.