Explore the vast range of titles available.

This paper presents the Edge Learning Machine (ELM), a machine learning framework for edge devices, which manages the training phase on a desktop computer and performs inferences on microcontrollers. The framework implements, in a platform-independent C language, three supervised machine learning algorithms (Support Vector Machine (SVM) with a linear kernel, k-Nearest Neighbors (K-NN), and Decision Tree (DT)), and exploits STM X-Cube-AI to implement Artificial Neural Networks (ANNs) on STM32 Nucleo boards. We investigated the performance of these algorithms on six embedded boards and six datasets (four classifications and two regression). Our analysis—which aims to plug a gap in the literature—shows that the target platforms allow us to achieve the same performance score as a desktop machine, with a similar time latency. ANN performs better than the other algorithms in most cases, with no difference among the target devices. We observed that increasing the depth of an NN improves performance, up to a saturation level. k-NN performs similarly to ANN and, in one case, even better, but requires all the training sets to be kept in the inference phase, posing a significant memory demand, which can be afforded only by high-end edge devices. DT performance has a larger variance across datasets. In general, several factors impact performance in different ways across datasets. This highlights the importance of a framework like ELM, which is able to train and compare different algorithms. To support the developer community, ELM is released on an open-source basis.

With the widespread use of electric bicycles and batteries, fire accidents caused by batteries have become increasingly serious, especially in closed indoor environments. Traditional fire prevention methods often rely on static monitoring and simple sensors, which can suffer from delayed responses or fail to accurately identify fire hazards. To address this issue, this paper proposes an innovative monitoring system for preventing fire hazards caused by batteries or electric bicycles in indoor environments, based on an STMicroelectronics 32-bit Microcontroller (STM32) and Open-source Machine Vision module (OpenMV). The system uses deep learning to train a neural network to recognize the image information of batteries or electric bicycles. The key innovation of this system lies in several aspects: firstly, it utilizes the OpenMV module for real-time image processing, enabling efficient and accurate recognition of batteries and electric bicycles; secondly, the integration with the STM32 microcontroller enhances the system’s data processing capabilities and enables flexible communication and responses with external devices; finally, the system features high-efficiency serial communication, ensuring the real-time transmission and processing of monitoring data for swift responses to potential fire risks. Experimental results show that the system can accurately identify batteries or electric bicycles in indoor environments and respond in a timely manner, significantly reducing fire hazards. In addition, the system's design is not limited to preventing battery-related fire hazards in indoor environments. Compared to traditional methods, this study's innovation lies in combining deep learning and embedded control technology for fire prevention, providing a practical and scalable solution for battery-related fire risk prevention.

The article describes the implementation of IoT technology in the teaching of microprocessor technology. The method presented in the article combines the reality and virtualization of the microprocessor technology laboratory. A created IoT monitoring device monitors the students’ microcontroller pins and sends the data to the server to which the teacher is connected via the control application. The teacher has the opportunity to monitor the development of tasks and student code of the program, where the functionality of these tasks can be verified. Thanks to the IoT remote laboratory implementation, students’ tasks during the lesson were improved. As many as 53% (n = 8) of those students who could improve their results achieved an improvement of one or up to two tasks during class. Before the IoT remote laboratory application, up to 30% (n = 6) of students could not solve any task and only 25% (n = 5) solved two tasks (full number of tasks) during the class. Before implementation, 45% (n = 9) solved one problem. After applying the IoT remote laboratory, these numbers increased significantly and up to 50% (n = 10) of students solved the full number of tasks. In contrast, only 10% (n = 2) of students did not solve any task.

The Chinese cooking process, intricate and diverse, encompasses a wide range of dishes, generating substantial data that often poses significant challenges for automation. To address these challenges, this paper introduces a versatile, highly reliable, user-friendly, and intelligent data recorder specifically designed for Chinese cuisine-the Chinese Cuisine Intelligent Cooker Recorder-V1.0 (CCICR-V1.0). This device is intended to record data generated during the cooking process and provide essential support for intelligent cooking robot systems. The core of CCICR-V1.0 is a STM32 microcontroller, equipped with sensors that monitor temperature, pressure, and attitude. These sensors facilitate the intelligent collection of data related to dishes, ingredients, and the movements of cookware throughout the cooking process. To ensure efficient data acquisition and storage, CCICR-V1.0 employs a multi-task data buffering storage method that effectively allocates CPU resources. NAND Flash is used as the storage medium, guaranteeing secure storage, management, and preservation of high-frequency, multi-channel data. Experimental results demonstrate the effectiveness of CCICR-V1.0 in achieving multi-channel, high-frequency data acquisition and storage in complex environments. This leads to an increase in automation efficiency of 89.9%. The weighing module demonstrates a maximum relative error of only 0.288%, while the attitude sensor experiment shows an attitude information error of just 0.22°C. Furthermore, the non-contact infrared temperature measurement module exhibits impressive performance with a maximum absolute error of 0.258 ∘ C and a maximum relative error of 0.88%. Notably, this measurement module has achieved a cost reduction of approximately 91.7% and an accuracy improvement of around 40%. Additionally, the response time of CCICR-V1.0 is less than 3 ms. With its ability to effectively capture data related to Chinese cooking, CCICR-V1.0 holds commercial value for widespread adoption and application in the field of intelligent cooking.

Accelerometers are nowadays included in almost any portable or mobile device, including smartphones, smartwatches, wrist-bands, and even smart rings. The data collected from them is therefore an ideal candidate to tackle human motion recognition, as it can easily and unobtrusively be acquired. In this work we analyze the performance of a hand-gesture classification system implemented using LSTM neural networks on a resource-constrained microcontroller platform, which required trade-offs between network accuracy and resource utilization. Using a publicly available dataset, which includes data for 20 different hand gestures recorded from 10 subjects using a wrist-worn device with a 3-axial accelerometer, we achieved nearly 90.25% accuracy while running the model on an STM32L4-series microcontroller, with an inference time of 418 ms for 4 s sequences, corresponding to an average CPU usage of about 10% for the recognition task.

In the increasingly developing logistics industry, the application of intelligent logistics cars is becoming more and more common. This paper mainly introduces a design scheme of intelligent logistics cars based on the STM32F103ZET6 single-chip microcomputer and OpenMV programmable camera. Its main process is as follows. First, through OpenMV, the car can scan a QR code to read relevant information of the task. Then, OpenMV recognizes the color of the object, and the single-chip microcomputer controls the manipulator arm to grab the object of the corresponding color. Next, the car will carry the object to the designated position along the designated route. Finally, the process is repeated several times until the delivery task is completed. The study of intelligent cars has very important practical value and theoretical significance.

Due to accelerated urbanization, modern urban residents are facing increasing life pressures. Many citizens are experiencing situational aversion in daily commuting, and the deterioration in the traffic environment has led to psychological distress of varying degrees among urban dwellers. Cyclists, who account for about 7% of urban commuters, lack a sense of belonging in the urban space and experience significant deficiencies in the corresponding urban infrastructure, which causes more people to face significant barriers to choosing cycling as a mode of transportation. To address the aforementioned issues, this study proposes a bionic intelligent interaction helmet (BIIH) designed and validated based on the principles of bionics, which has undergone morphological design and structural validation. Constructed around the STM32-embedded development board, the BIIH is an integrated smart cycling helmet engineered to perceive environmental conditions and enable both human–machine interactions and environment–machine interactions. The system incorporates an array of sophisticated electronic components, including temperature and humidity sensors; ultrasonic sensors; ambient light sensors; voice recognition modules; cooling fans; LED indicators; and OLED displays. Additionally, the device is equipped with a mobile power supply, enhancing its portability and ensuring operational efficacy under dynamic conditions. Compared with conventional helmets designed for analogous purposes, the BIIH offers four distinct advantages. Firstly, it enhances the wearer’s environmental perception, thereby improving safety during operation. Secondly, it incorporates a real-time interaction function that optimizes the cycling experience while mitigating psychological stress. Thirdly, validated through bionic design principles, the BIIH exhibits increased specific stiffness, enhancing its structural integrity. Finally, the device’s integrated power and storage capabilities render it portable, autonomous, and adaptable, facilitating iterative improvements and fostering self-sustained development. Collectively, these features establish the BIIH as a methodological and technical foundation for exploring novel research scenarios and prospective applications.

The advantageous applications of magnetic bistable microwires have emerged during long-lasting research. They have a wide range of applications in the scientific sphere or technical practice. They can be used for various applications, including magnetic memories, biomedicine, and sensors. This manuscript is focused on the last-mentioned application of microwires—sensors—discussing various digital signal processing techniques used in practical applications. Thanks to the highly sensitive properties of microwires and their two stable states of magnetization, it is possible to perform precise measurements with less demanding digital processing. The manuscript presents four practical signal-processing methods of microwire response using three different experiments. These experiments are focused on detecting the signal in a simple environment without an external magnetic background, measuring with the external background of a ferromagnetic core, and measuring in harsh conditions with a strong magnetic background. The experiments aim to propose the best method under various conditions, emphasizing the quality and signal processing speed of the microwire signal.

During takeoff and landing, birds bounce and grab with their legs and feet. In this paper,the lower limb structure of the bionic bird is designed with reference to the function of jumping and grasping, and the PID algorithm based on the development module of stm32 development board is used to speed control the lower limb driving element, so that the motor and the bishaft steering gear move with the rate change of sine wave. According to the speed of grasping response time and the size of grasping force, the structure of the bionic bird paw is designed. Based on the photosensitive sensor fixed in the geometric center of the foot, the grasping action of the lower limb mechanism is intelligently controlled. Finally, the kinematic verification of the lower limb structure is carried out by ADAMS. Experiments show that the foot structure with four toes and three toes is more conducive to maintaining the stability of the body while realizing the fast grasping function. In addition, it can effectively improve the push-lift ratio of the bionic ornithopter by adjusting the sinusoidal waveform rate of the motor speed.

This paper addresses the low level of intelligence in tea processing equipment in Enshi Prefecture by designing an intelligent withering control system based on the STMicroelectronics 32-bit Microcontroller (STM32). This control system can achieve real-time monitoring of the withering environment and automate the control of heating and ventilation dehumidification modules. By integrating IoT technology, relevant users can view the tea production process via mobile devices, enabling intelligent and remote production operations. Application results show that the system operates stably, accurately measures temperature and humidity in the withering environment, and achieves a control precision of % through a fuzzy control algorithm. It effectively meets the needs of tea processing in Enshi Prefecture. The system not only optimizes traditional processing workflows, enhancing processing efficiency and tea quality, but also provides new technological means for tea processing enterprises, contributing to the development and upgrading of the local tea industry.