Minimizing the Cost for Resource Allocation from Multiple Cloud Providers

Cloud computing provides streamlined tools for exceptional business efficiency. Cloud service providers typically offer two types of plans: reserved and on-demand. Restricted policies provide low-cost long-term contracting, while order contracts are expensive and ready for short periods. Cloud resources must be delivered wisely to meet current customer demands. Many current works rely on low-cost resource-reserved strategies, which may be under- or over-provisioning. Resource allocation has become a difficult issue due to unfairness causing high availability costs and cloud demand variability. That article suggests a hybrid approach to allocating cloud services to complex customer orders. The strategy was built in two stages: accommodation stages and a flexible structure. By treating each step as an optimization problem, we can reduce the overall implementation cost while maintaining service quality. Due to the uncertain nature of cloud requests, we set up a stochastic Optimization-based approach. Our technique is used to assign individual cloud resources and the results show its effectiveness. Keywords: Cloud computing, Resource allocation, Demand

Empirical Assessment of Bullwhip Effect in Supply Networks

Previous empirical studies on bullwhip effects treat each industry or firm as isolated from its supply chain network. In this paper, the authors are interested in the role played by supply chain relational connection in moderating how demand variability signal is transmitted upstream. The paper conducts an empirical study based on a panel data of 55 manufacturing industries and 9 wholesale industries. The regression analysis shows that demand variability is propagated through supply chain upward and the transmission is influenced by the structural relationship between suppliers and customers, which is measured by customer-base concentration and customer interconnectedness. On the other hand, customer demand variability has a greater impact on industries with less concentrated customer base or with less interconnected customers.

Optimal Design of District Metered Areas in a Water Distribution Network Using Coupled Self-Organizing Map and Community Structure Algorithm

Operation and management of a water distribution network (WDN) by district metered areas (DMAs) bring many benefits for water utilities, particularly regarding water loss control and pressure management. However, the optimal design of DMAs in a WDN is a challenging task. This paper proposes an approach for the optimal design of DMAs in the multiple-criteria decision analysis (MCDA) framework based on the outcome of a coupled model comprising a self-organizing map (SOM) and a community structure algorithm (CSA). First, the clustering principle of the SOM algorithm is applied to construct initial homologous clusters in terms of pressure and elevation. CSA is then coupled to refine the SOM-based initial clusters for the automated creation of multiscale and dynamic DMA layouts. Finally, the criteria for quantifying the performance of each DMA layout solution are assessed in the MCDA framework. Verifying the model on a hypothetical network and an actual WDN proved that it could efficiently create homologous and dynamic DMA layouts capable of adapting to water demand variability.

Improvising Dynamic Cloud Resource Allocation to Optimise QoS and Cost Effectiveness

Cloud computing offers streamlined instruments for outstanding business efficiency processes. Cloud distributors typically give two distinct forms of usage plans: Reserved as well as On-demand. Restricted policies provide inexpensive long-term contracting services, while order contracts were very expensive and ready for brief rather than long longer periods. In order to satisfy current customer demands with equal rates, cloud resources must be delivered wisely. Many current works depend mainly on low-cost resource-reserved strategies, which may be under-provisioning and over-provisioning rather than costly ondemand solutions. Since unfairness can cause enormous high availability costs and cloud demand variability in the distribution of cloud resources, resource allocation has become an extremely challenging issue. The hybrid approach to allocating cloud services according to complex customer orders is suggested in that article. The strategy was constructed as a two-step mechanism consisting of accommodation stages and then a versatile structure. In this way, by constructing each step primarily as an optimization problem, we minimize the total cost of implementation, thereby preserving service quality. By modeling client prerequisites as probability distributions are disseminated owing to the dubious presence of cloud requests, we set up a stochastic Optimization-based approach. Using various approaches, our technique is applied, and the results demonstrate its effectiveness when assigning individual cloud resources.

The Bullwhip Effect in Supply Networks

We offer a new network perspective on one of the central topics in operations management—the bullwhip effect (BWE). The topic has both practical and scholarly implications. We start with an observation: the variability of orders placed to suppliers is larger than the variability of sales to customers for most firms, yet the aggregate demand variability felt by suppliers upstream does not amplify commensurably. We hypothesize that changes to the supplier’s customer base can smooth out its aggregate demand. We test the hypothesis with a data set that tracks the evolution of supply relationships over time. We show that the effect of customer base management extends beyond the statistical benefits of aggregation. In particular, both the formation and the dissolution of customer-supplier relationships are associated with the smoothing of the aggregate demand experienced by suppliers. This provides fresh insight into how firms may leverage their customer-supplier relationships to mitigate the impact of the BWE. This paper was accepted by Jay Swaminathan, operations management.

A two-stage stochastic optimization model for the retail multiskilled personnel scheduling problem: a k-chaining policy with $k \ge 2$

<abstract> <p>Considering an uncertain demand, this study evaluates the potential benefits of using a multiskilled workforce through a k-chaining policy with $k \ge 2$. For the service sector and, particularly for the retail industry, we initially propose a deterministic mixed-integer linear programming model that determines how many employees should be multiskilled, in which and how many departments they should be trained, and how their weekly working hours will be assigned. Then, the deterministic model is reformulated using a two-stage stochastic optimization (TSSO) model to explicitly incorporate the uncertain personnel demand. The methodology is tested for a case study using real and simulated data derived from a Chilean retail store. We also compare the TSSO approach solutions with the myopic approaches' solutions (i.e., zero and total multiskilling). The case study is oriented to answer two key questions: how much multiskilling to add and how to add it. Results show that TSSO approach solutions always report maximum reliability for all levels of demand variability considered. It was also observed that, for high levels of demand variability, a k-chaining policy with $k \ge 2$ is more cost-effective than a 2-chaining policy. Finally, to evaluate the conservatism level in the solutions reported by the TSSO approach, two truncation types in the probability density function (pdf) associated with the personnel demand were considered. Results show that, if the pdf is only truncated at zero (more conservative truncation) the levels of required multiskilling are higher than when the pdf is truncated at 5th and 95th percentiles (less conservative truncation).</p> </abstract>

São Felipe Hospital: Capacity Management Challenge in a Hospital Emergency

ABSTRACT In July 2017, the board of directors of Hospital São Felipe, a traditional hospital located in Minas Gerais, met to discuss the results of the satisfaction survey conducted at the hospital, where it was clear there was great customer dissatisfaction with the emergency service. In the previous year, the hospital emergency service received, on average, about 6,300 patients a month, divided in three specialties: general clinic, orthopedics, and ophthalmology. The director of emergency services had twenty days to submit a plan of action to address the problems identified in the emergency area, particularly those related to the waiting lines: wait time, lack of comfort, inattention of employees, and so on. The first action taken by the director was to collect data that would enable him to analyze wait times during the process: What time did the patient arrive at the emergency service? How long the patient waited to be attended by the receptionist? How long the patient waited for triage? and so on. With these data, he believed that he would have a better understanding of the process flow and would be able to propose solutions to the problem of waiting lines in the emergency area. The case was written with the educational goal of working with the concept of capacity management in services and with ways to deal with the demand variability, especially in high-touch and unpredictable services, as in the case of an emergency service.

Hospital São Felipe: Desafio da Gestão da Capacidade em uma Emergência Hospitalar

ABSTRACT In July 2017, the board of directors of Hospital São Felipe, a traditional hospital located in Minas Gerais, met to discuss the results of the satisfaction survey conducted at the hospital, where it was clear there was great customer dissatisfaction with the emergency service. In the previous year, the hospital emergency service received, on average, about 6,300 patients a month, divided in three specialties: general clinic, orthopedics, and ophthalmology. The director of emergency services had twenty days to submit a plan of action to address the problems identified in the emergency area, particularly those related to the waiting lines: wait time, lack of comfort, inattention of employees, and so on. The first action taken by the director was to collect data that would enable him to analyze wait times during the process: What time did the patient arrive at the emergency service? How long the patient waited to be attended by the receptionist? How long the patient waited for triage? and so on. With these data, he believed that he would have a better understanding of the process flow and would be able to propose solutions to the problem of waiting lines in the emergency area. The case was written with the educational goal of working with the concept of capacity management in services and with ways to deal with the demand variability, especially in high-touch and unpredictable services, as in the case of an emergency service.

Consumer Learning from Own Experience and Social Information: An Experimental Study

We investigate how different types of social information affect the demand characteristics of firms competing through service quality. We first generate behavioral hypotheses around both consumers’ learning behavior and firms’ corresponding demand characteristics: market share, demand uncertainty, and rate of convergence. We then conduct a controlled human-subject experiment in which a consumer chooses to visit one of two firms, each with unknown service quality, in a repeated interaction and is exposed to different information treatments from a social network: (1) no social information; (2) share-based social information, which details the percentage of people who visited each firm; (3) quality-based social information, which illustrates the percentage of people who received a satisfactory experience from each firm; or (4) full social information, which contains both share- and quality-based social information. A key insight from our study is that different types of social information have different effects on firms’ demand. First, promoting quality-based social information leads to a significantly higher market share, lower demand variability, and faster rate of convergence for a firm with significantly better service quality. Second, when the higher quality firm has only a marginal advantage over the other firm, promoting only share-based information leads to significantly higher market share and lower demand variability. A third important result is that providing only one type of social information can actually be more helpful to the higher quality firm than providing full social information. This paper was accepted by David Simchi-Levi, operations management.