Sistemas de prevención activa de riesgos laborales para maquinaria industrial

Abstract

The results obtained in this Ph.D. thesis open a new research line, focused on the design and development of Safety Measures for the elimination and reduction of risks associated with industrial machinery, in an active fashion, i. e., automatically (without being activated by a worker) and with minimum delay. Currently, Electro-sensitive Protective Equipment (ESPEs) are the only type of Safety Measures capable of performing active occupational hazard prevention. An ESPE comprises three parts: one or more sensors (e.g., photoelectric, laser), an output switch (OSSD) connected to the machine control system to prevent or allow its functions, and a control unit that manages the information provided by the sensor and the behavior of the OSSD. Vision Based Protective Devices (VBPDs) are most recently developed group of ESPEs. VBPDs are designed to instantly detect access of objects to the danger zone of a machine, from visual information captured by a Computer Vision System (CVS). A VBPD impedes machine functions by the automatic actuation of its OSSD when it detects and object accessing to that zone. The work performed in this Ph.D. thesis has focused on the design and development VBPDs, as well as on the analysis and optimization of algorithms suitable for their CVS. For this purpose, several tests have been carried out, providing a safe means to control VBPD behaviour under a series of configurable parameters, thus enabling rapid prototyping and easy adaptation to existing machinery applications.