Many of the existing Leader Election algorithms don't deal with energy consumption and fault tolerance since they are not mainly dedicated to autonomous systems like wireless sensor and IoT networks. It is possible to use the classical Minimum Finding (MinFind) algorithm, where each node sends its value in a broadcast mode each time a better value is received. This process is very energy consuming and not realistic since it may be subject to an important number of collisions. In this paper, we propose a new algorithm called WBS (Wait-Before-Starting), which is fault tolerant and where each node in the network will wait for a certain time before starting the execution of its program. This time lapse depends on the node's value so that the leader will be the one which will wait the least. In this case, the leader will be the first one which will start by sending a message to all the nodes of the network to inform them that it is the leader. Then, the other nodes will start the execution of their programs. Otherwise, if the leader fails, another node will do the same. The obtained results show that the proposed algorithm reduces the energy consumption with rates that can exceed 96% compared with the classical MinFind Algorithm.