Search

We use discrete event simulation to model and analyze a real-life emergency department (ED). Our approach relies on the appropriate integration of most real-life ED features to the simulation model in order to derive useful practical results. Data is supplied from the ED of the urban French hospital Saint Camille. Our purpose is to optimize the human resource staffing levels. We want to minimize the patient average length of stay ( LOS¯), by integrating the staffing budget constraint and a constraint securing that the most severe incidents will see a doctor within a specified time limit. The second constraint allows to avoid the perverse effect of only considering the LOS¯ metric that would delay the treatment of the most urgent patients. We use simulation-based optimization, in which we perform a sensitivity analysis expressing LOS¯ as a function of the staffing budget and also the average door-to-doctor time for urgent patients ( DTDT¯). We show that the budget has a diminishing marginal effect on the problem solution. Due to the correlation between LOS¯ and DTDT¯, we also observe that the DTDT¯ constraint may significantly affect the feasibility of the problem or the value of the optimal solution. [ABSTRACT FROM AUTHOR]

In the context of multi-channel call centers with inbound calls and emails, this article considers a threshold policy on the reservation of agents for the inbound calls. We study a general non-stationary model where calls arrive according to a non-homogeneous Poisson process. The optimization problem consists in maximizing the throughput of emails under a constraint on the waiting time of inbound calls. An efficient adaptive threshold policy is proposed that is easy to implement in the automatic call distributor. This scheduling policy is evaluated through a comparison with the optimal performance measures found in the case of a constant arrival rate and also with other intuitive adaptive threshold policies in the general non-stationary case. [ABSTRACT FROM PUBLISHER]

We study call routing policies for call centers with multiple call types and multiple agent groups. We introduce new weight-based routing policies where each pair (call type, agent group) is given a matching priority defined as an affine combination of the longest waiting time for that call type and the longest idle time or the number of idle agents in that agent group. The coefficients in this combination are parameters to be optimized. This type of policy is more flexible than traditional ones found in practice, and it performs better in many situations. We consider objective functions that account for the service levels, the abandonment ratios, and the fairness of occupancy across agent groups. We select the parameters of all considered policies via simulation-based optimization heuristics. This requires only the availability of a simulation model of the call center, which can be much more detailed and realistic than the models used elsewhere in the literature to study the optimality of certain types of routing rules. We offer a first numerical study of realistic routing rules that takes into account the complexity of real-life call centers. [ABSTRACT FROM AUTHOR]

In this paper, we discuss the dynamic server control in a two-class service system with abandonments. Two models are considered. In the first case, rewards are received upon service completion, and there are no abandonment costs (other than the lost opportunity to gain rewards). In the second, holding costs per customer per unit time are accrued, and each abandonment involves a fixed cost. Both cases are considered under the discounted or average reward/cost criterion. These are extensions of the classic scheduling question (without abandonments) where it is well known that simple priority rules hold. The contributions in this paper are twofold. First, we show that the classic c- μ rule does not hold in general. An added condition on the ordering of the abandonment rates is sufficient to recover the priority rule. Counterexamples show that this condition is not necessary, but when it is violated, significant loss can occur. In the reward case, we show that the decision involves an intuitive tradeoff between getting more rewards and avoiding idling. Secondly, we note that traditional solution techniques are not directly applicable. Since customers may leave in between services, an interchange argument cannot be applied. Since the abandonment rates are unbounded we cannot apply uniformization-and thus cannot use the usual discrete-time Markov decision process techniques. After formulating the problem as a continuous-time Markov decision process (CTMDP), we use sample path arguments in the reward case and a savvy use of truncation in the holding cost case to yield the results. As far as we know, this is the first time that either have been used in conjunction with the CTMDP to show structure in a queueing control problem. The insights made in each model are supported by a detailed numerical study. [ABSTRACT FROM AUTHOR]

This paper focuses on monotonicity results for dynamic systems that take values in the natural numbers or in more-dimensional lattices. The results are mostly formulated in terms of controlled queueing systems, but there are also applications to maintenance systems, revenue management, and so forth. We concentrate on results that are obtained by inductively proving properties of the dynamic programming value function. We give a framework for using this method that unifies results obtained for different models. We also give a comprehensive overview of the results that can be obtained through it, in which we discuss not only (partial) characterizations of optimal policies but also applications of monotonicity to optimization problems and the comparison of systems. [ABSTRACT FROM AUTHOR]

We study a shift scheduling problem for call centers with an overall service level objective. We prove a property of this problem, called multimodularity, that ensures that a local search algorithm terminates with a globally optimal solution. We report on computations performed using real call center data. [ABSTRACT FROM AUTHOR]