Cyber-Physical systems (CPS) are the new generation of intelligent systems, which gives greater importance to the interaction between the physical processes and the computational components than their predecessors. It has become one of the fundamental pillars of the fourth industrial revolution which has brought with it the emergence of new challenges in the design of their control systems. This paper deals with the basic concepts of CPS and industrial cyber-physical systems (ICPS), presents a review of the main control strategies used for its design.

DFA (Design for assembly) is an important part of the contents included in some of the manufacturing courses taught at Jaume I University. DFA is a tool to analyse and improve product design from an assembly point of view. Although DFA contents had been covered using different teaching activities (theoretical, problem and laboratory sessions), the results in students’ assessment revealed that the expected learning outcomes were not being achieved. In particular, results were especially unsatisfactory in the practical application of DFA. Students misunderstood concepts such as “handling” and “insertion” operations, and failed at identifying assembly problems related to thickness or alignment among others. A learning by doing approach has been proved to improve students’ learning and engagement, as they take an active role and have the opportunity of doing things themselves. In a previous work, a specific modular and reconfigurable kit to improve DFA learning by experimentation was designed. Based on this work, this paper presents the analysis of the results obtained in two different courses where the modular kit has been used by students in a new seminar session.

En este artículo se presenta el desarrollo e implementación de una estrategia de supervisión de un proceso de microfabricación. El método propuesto está basado en técnicas de Inteligencia Artificial, embebidas en una plataforma de tiempo real para la monitorización inteligente de procesos. La contribución se centra fundamentalmente en dos modelos para la estimación en proceso (on-line) de la rugosidad superficial (Ra), a partir de la mínima información sensorial posible. El primero de estos modelos está basado en un algoritmo para el modelado híbrido incremental (HIM), cuyos parámetros óptimos se obtienen a partir de un método estocástico, representado por el temple simulado. El segundo está basado en un algoritmo de agrupamiento borroso generalizado (GFCM), incorporado en un sistema de inferencia de una estructura neuroborrosa. Esta estrategia se embebe en una plataforma para una ejecución en tiempo real y en paralelo junto con el resto de estrategias y métodos. Finalmente, se hace una validación en una plataforma experimental, utilizada como soporte tecnológico, lo cual permite el aprovechamiento mutuo de las experiencias alcanzadas y la mejora de los resultados obtenidos. Este resultado científico y técnico, supone un salto cualitativo importante sin precedentes en la investigación industrial en el campo de la microfabricación.

No abstract