Um sistema evolutivo para alocação de serviços de acordo com o perfil dos funcionários

Abstract

The problem of service allocation according to the profile of employees that belonging to a particular division of a company seems a trivial task, but if analyzed more cautiously it presents some conflicting objectives and constraints, which makes it very complex. To accomplish this allocation should be considered: the maximum number of employees in the division, the staff heterogeneity of training and experience, if there is enough material to accomplish the job, the time needed do finish the service, the employee’s idle time, etc. There are on the market some systems used by the industry, they propose to allocate services using both computational approaches the conventional ones (using graphs and mathematical programming) an the intelligent ones (search algorithms, neural networks and genetic algorithms). But regardless of the adopted approach this systems do not take into account the allocated unit intrinsic characteristics for each service, to optimize the realized work. This paper presents the specification of a system based on evolutionary computation, more specifically on a variation of Holland’s Genetic Algorithm, to assist the maintenance section of the Coordenação de Gestão e Tecnologia da Universidade Federal de Alagoas, and can be expanded to all the university, in the task to designate a skilled professional to fix a problem presented by a computer owned by the Instituto de Ciências Humanas, Comunicação e Artes da Universidade Federal de Alagoas.