This paper presents a first approach to a new QoS aware routing suitable for wireless and wired ad-hoc networks. The proposed algorithm is based on artificial intelligent techniques. Actually, it is a merge between a genetic algorithm and the TBR (ticket based routing) algorithm. A better TBR performance is pursued in order to use the result in ubiquitous sensor access networks. It has been necessary to choose a routing protocol as a starting point and to define how artificial intelligent techniques will be applied in order to improve the overall performance. The resulting algorithm is an evolution of TBR, an on-demand routing protocol for ad-hoc networks that provides quality of service. The performance of TBR has been improved by reducing the load of routing packets in the network and by minimizing the communication latency due to its on-demand behavior. Moreover, with the genetic algorithm, the likelihood of finding the route optimum using several metrics has been increased. The main goal is the implementation of an approach that will make possible a future formal specification of evolutionary routing algorithm for USN. In this paper, first simulation results are presented. Moreover, the chromosome coding, crossover and mutation operators and the fitness evaluation formula are described.