AN OBSTACLE-AWARE MODEL TO INDOOR NAVIGATION APPLIED TO SUPERMARKET SHOPPING ROUTING
Wagner Cipriano da Silva1; Paulo Eduardo Maciel de Almeida1
1 CEFET-MG - Centro Federal de Educação Tecnológica de Minas Gerais
doi:10.20906/CPS/CILAMCE2017-1091
Resumo
Some smart applications make sense only when the provider knows exactly where the customer is, so indoor localization systems are important to support smart services when outdoor positioning is not possible. Nowadays, the main way to define position over the Earth is the GPS, that is useful for outdoor location positioning, but has strong limitations for indoor services. Urban population expansion brings challenges and demands public places with adequate structure. Accompanying this development, retail establishments grow gradually in scale and in number of products. Then, the customers need services which could find good routes from actual location to points of interest, for instance, to reduce shopping time. This paper proposes an optimization model that uses genetic algorithms (GA) to solve a graph-based routing model. Hierarchical graphs are used to get the indoor map of a retail establishment, and to represent the commodities that make up the shopping list. Then, GA meta-heuristics are adopted to bring out short routes for the shopping cart. The results show that this model is flexible for different retail store layouts and can find good solutions in reasonable time for the supermarket shopping routing problem.
Palavras-chave: Computational intelligence; optimization; shopping route planning; indoor navigation