La soldadura por puntos de resistencia es una técnica común en la industria manufacturera, especialmente en el sector automotriz en México, debido a su versatilidad y facilidad de implementación. La evaluación de la calidad de estas soldaduras es crucial, y existen múltiples métodos para ello, lo que complica la elección del más adecuado. Este artículo presenta una revisión de la literatura centrada en métodos no destructivos para la confirmación de la calidad de la soldadura por puntos de resistencia. Se destacan técnicas como la inspección por ultrasonido, la termografía infrarroja y la visión por computadora, entre otras. La metodología empleada incluye un análisis comparativo de estudios recientes para identificar las mejores prácticas y aplicaciones en la evaluación de la calidad de la soldadura por puntos. Los resultados de esta revisión proporcionan una guía comprensiva para investigadores y profesionales, facilitando la selección de métodos adecuados para aplicaciones específicas.

Según un informe de la IEA de 2017, el consumo de energía en los edificios es aproximadamente un tercio del consumo total de energía en el mundo. Además, la iluminación representa el 19% del consumo de electricidad en todo el mundo. Tomando como referencia estos datos, surge la necesidad de implementar una solución que permita reducir dichos consumos. Una forma de hacerlo es la implementación de dispositivos que permitan convertir las áreas de trabajo en espacios inteligentes. El objetivo de este trabajo es la implementación de una oficina inteligente, monitoreando y controlando el consumo energético de los dispositivos eléctricos mediante tecnologías IoT, logrando un ahorro energético significativo y a la vez proporcionando confort al personal que labora en dicha oficina. Para ello, se instalaron diversos sensores tales como sensores para la detección del estado de las puertas (abiertas o cerradas), sensores de presencia (para determinar si se encentran o no personas en áreas específicas) y actuadores, los cuales, con base a las señales transmitidas por los sensores instalados, se pudieron crear escenas que permiten el control de la iluminación de manera automática. La Raspberry Pi 4 se utilizó como controlador central, se integró la herramienta Node-Red para establecer la comunicación entre los dispositivos IoT. Node-Red es una herramienta de desarrollo basada en flujo para la programación visual que permite conectar dispositivos de hardware, API y servicios en línea. Esta herramienta nos facilita la integración de diferentes sensores y marcas, en donde se realizó la programación de diferentes escenas para poder generar un ahorro energético de manera automática. Asimismo, se estableció comunicación con una nube para poder monitorear y controlar de manera remota el sistema implementado mediante el protocolo de comunicación MQTT, el cual es un protocolo de transmisión de datos diseñado para la transición de datos de máquina a máquina. Un servidor, denominado “bróker” de mensajes, recibe transmisiones de dispositivos que “publican” información en “tópicos” particulares y, posteriormente, entregan esos mensajes a otras máquinas que se “suscriben” a esos “tópicos” específicos. Para lograr esto, se utilizaron los servicios de HiveMQ, los cuales permiten conectar hasta 100 dispositivos de manera gratuita. El resultado del proyecto es un sistema que puede ser controlado de manera remota que, debido a las tecnologías implementadas, ha genera un ahorro energético del 6.1%. Este ahorro reduce los costos operativos y apoya los objetivos de sostenibilidad de la oficina. Además, el sistema brinda confort para las personas que trabajan en dicha oficina. También cabe mencionar que este proyecto puede ser escalable, es decir, a medida que la oficina amplíe sus instalaciones, el sistema puede ser modificado para agregar los sensores y actuadores que se requieran para poder controlar dicha ampliación.

Edge detection is a key building block in real-time image-processing applications such as drone-based infrastructure inspection, autonomous navigation, and remote sensing. However, its computational cost remains a challenge for resource-constrained embedded systems. This work presents a hardware-accelerated edge detection architecture based on a homogeneous 2D-mesh Network-on-Chip (NoC) integrating systolic arrays to efficiently perform the convolution operations required by the Sobel filter. The proposed architecture was first developed and validated as a 3 × 3 mesh prototype on FPGA (Xilinx Zynq-7000, Zynq-7010, XC7Z010-CLG400A, Zybo board, utilizing 26,112 LUTs, 24,851 flip-flops, and 162 DSP blocks), achieving a throughput of 8.8 Gb/s with a power consumption of 0.79 W at 100 MHz. Building upon this validated prototype, a reduced 2 × 2 node cluster with 14-bit word width was subsequently synthesized at the physical level as a proof-of-concept using the OpenLane RTL-to-GDSII open-source flow targeting the SkyWater 130 nm PDK (sky130A). Post-layout analysis confirms the manufacturability of the design, with a total power consumption of 378 mW and compliance with timing constraints, demonstrating the feasibility of mapping the proposed architecture to silicon and its suitability for drone-based infrastructure monitoring applications.

This work was originated from the increasing interest in several industries to implement voice based virtual assistant solutions powered by the Natural Language Processing field of study. This work is focused on the automotive industry Human Machine Interface related products, specifically the Instrument Panel. Nowadays people are constantly using virtual assistants like Google Assistant, Alexa, Cortana or Siri on their electronic devices. Furthermore, 31% of cars have a built-in virtual assistant, for example Ford uses Alexa, Merced­es-Benz and Hyundai use Google Assistant, BMW and Nissan use Cortana, GM uses IBM Watson, Honda uses Hana and Toyota uses YUI. Apart from the proprietary solutions described earlier, there are also contemporary open-source generic solutions available on the market, such as Mycroft AI which stands out from other technologies due to ready to deploy, well documented, simple installation on a Linux PC or RPI SoC, and simple execution. This paper presents a way to use Mycroft AI as an alternative to add artificial intelligence-based voice assistance to applications in the automotive domain. The voice communication module presented here drives notifications related to three different entities: seat belt, fuel level and battery level, all of them are telltales present in any automotive Instrument Panel. Since the Mycroft AI design approach is based on Human Centered Design (HCD), the voice communication module presented here provides real user experience (UX) based design. As a conclusion, Mycroft AI demonstrates great potential as an alternative to add voice assistance to automotive industry Human Machine Interface related products. About future work, due to the fact that Mycroft AI is based on Python, there are many possibilities for connecting and expanding the voice communication module by using countless Python libraries in order to import and process any type of information, in any format or source, for example the information from communication technologies like CAN, LIN, Ethernet, MOST, GPS or any other device or technology in order to create comprehensive automotive solutions.

Organic electrochemical transistors (OECTs) based on poly(3,4ethylenedioxythiophene) doped with poly (styrene sulfonate) (PEDOT:PSS) semiconductor have emerged as promising candidates for neuromorphic devices due to their mixed ionic–electronic transport of this polymer and their ability to emulate biological synaptic behavior. In this work, we report on the fabrication and characterization of flexible thick-film PEDOT:PSS OECTs produced by aerosol jet printing to explore the relationship between drain bias and memory performance in thick PEDOT:PSS channels. We found that these devices exhibited both short-term and long-term memory dynamics, depending on the applied gate-pulse number and pulse duration. We found that thick PEDOT:PSS channels displayed enhanced hysteresis in transfer curves, consistent with a larger ionic reservoir that allows prolonged charge trapping and a slower dedoping. A systematic study of the drain voltage ( V _D ) revealed that higher | V _D | values enhance the memory level by increasing both the current modulation and transconductance, as a larger fraction of the PEDOT:PSS channel participates in the electrochemical redox process, although the conventional voltage memory window exhibits a non-monotonic dependence on drain bias. Instead of relying exclusively on the conventional voltage memory window, an analysis of the effective memory window, which was defined as the enclosed area of the forward and backward transfer characteristics, was proposed to evaluate the effective ionic memory behavior of the OECTs, finding a linear trend with respect to V _D . These findings highlight the effect of drain bias in thick PEDOT:PSS channels offering design guidelines for future bioinspired, and memory-based printed and flexible organic electronics.

This work presents the proposal of an automatic method for monitoring fault codes of industrial processes, as well as its real implementation in a production line made up of 5 different processes. For the implementation, some industry 4.0 techniques were taken as a basis, such as Big Data, Industrial Networks, Manufacturing Execution Systems and Cloud Computing, integrating them into a system capable of collecting information in real time and focused on a automatic failure monitoring of the process in which it is implemented. A method is presented that consists of 12 specific steps, defined and applicable to a system of microcontrollers that guide processes where it is necessary to track the fault codes of the same process, and which are connected by some means of communication to share information. In the implementation of the method in a real case, you can see how to adapt it to manufacturing processes with great accuracy. This implementation was carried out in a serial production line made up of 5 processes dependent on each other where each process has its own controller and its own fault codes. The results of the implementation made visible the causes of the large number of process stoppages and the data returned by the system helped make important decisions to allocate resources efficiently in order to reduce the time lost due to machinery failures. It was also possible to locate the processes with the most problems in the last period, which made it possible to identify an area of ​​opportunity to increase the productivity of the monitored process.