Advanced sensing strategies in the industrial sector are becoming a valued technological answer to increase the performance and competitiveness. The development of enhanced sensing solutions considering both technology and monitoring requirements is, nowadays, subject of concern in the industrial maintenance field. In this context, this paper presents a novel self-powered wireless sensor applied to condition monitoring of gears. The proposed sensor is based on a modular architecture, offering multipoint sensing, local wireless communication, multisource energy harvesting, and embedded diagnosis algorithm for mechanical fractures detection based on acoustic emission analysis. The developments are complemented by means of a remote management interface, from which the user can configure the functionalities of the sensors, visualize the network status as well as analyze the diagnosis evolution. The sensor performance, in terms of power consumption and fault detection, has been analyzed by means of experimental results.