Esca is one of the most common grape leaf diseases that seriously affect grape yield, causing a loss of global production in the range of 20%-40%. Therefore, a timely and effective identification of the disease could help to develop an early treatment approach to control its spread while reducing economic losses. For this purpose the use of computer vision and machine learning techniques for recognizing plant diseases have been extensively studied in recent years. The aim of this paper is to propose an image detector based on a high-performance convolutional neural network (CNN) implemented in a low cost, low power platform, to monitor the Esca disease in real-time. To meet the severe constraints typical of an embedded system, a new low-rank CNN architecture (LR-Net) based on CANDECOMP/PARAFAC (CP) tensor decomposition has been developed. The compressed CNN network so obtained has been trained on a specific dataset and implemented in a low-power, low-cost Python programmable machine vision camera for real-time classification. An extensive experimentation has been conducted and the results achieved show the superiority of LR-Net with respect to the state-of-the-art networks both in terms of inference time and memory occupancy.

In this work, a new custom design of an anomaly detection and classification system is proposed. It is composed of a convolutional Auto-Encoder (AE) hardware design to perform anomaly detection which cooperates with a mixed HW/SW Convolutional Neural Network (CNN) to perform the classification of detected anomalies. The AE features a partial binarization, so that the weights are binarized while the activations, associated to some selected layers, are non-binarized. This has been necessary to meet the severe area and energy constraints that allow it to be integrated on the same die as the MEMS sensors for which it serves as a neural accelerator. The CNN shares the feature extraction module with the AE, whereas a SW classifier is triggered by the AE when a fault is detected, working asynchronously to it. The AE has been mapped on a Xilinx Artix-7 FPGA, featuring an Output Data Rate (ODR) of 365 kHz and achieving a power dissipation of $333 \boldsymbol {\mu } $ W/MHz. Logic synthesis has targeted TSMC CMOS 65 nm, 90 nm, and 130 nm standard cells. Best results achieved highlight a power consumption of $138 \boldsymbol {\mu } $ W/MHz with an area occupation of 0.49 mm ² when real-time operations are set. These results enable the integration of the complete neural accelerator in the CMOS circuitry that typically sits with the inertial MEMS on the same silicon die. Comparisons with the related works suggest that the proposed system is capable of state-of-the-art performances and accuracy.

Electric mobility is pervasive and strongly affects everyone in everyday life. Motorbikes, bikes, cars, humanoid robots, etc., feature specific battery architectures composed of several lithium nickel oxide cells. Some of them are connected in series and others in parallel within custom architectures. They need to be controlled against over current, temperature, inner pressure and voltage, and their charge/discharge needs to be continuously monitored and balanced among the cells. Such a battery management system exhibits embarrassingly parallel computing, as hundreds of cells offer the opportunity for scalable and decentralized monitoring and control. In recent years, tiny machine learning has emerged as a data-driven black-box approach to address application problems at the edge by using very limited energy, computational and storage resources to achieve under mW power consumption. Examples of tiny devices at the edge include microcontrollers capable of 10–100 s MHz with 100 s KiB to few MB embedded memory. This study addressed battery management systems with a particular focus on state-of-charge prediction. Several machine learning workloads were studied by using IEEE open-source datasets to profile their accuracy. Moreover, their deployability on a range of microcontrollers was studied, and their memory footprints were reported in a very detailed manner. Finally, computational requirements were proposed with respect to the parallel nature of the battery system architecture, suggesting a per cell and per module tiny, decentralized artificial intelligence system architecture.

Tiny machine learning technologies are bringing intelligence ever closer to the sensor, thus enabling the key benefits of edge computing (e.g., reduced latency, improved data security, higher energy efficiency, and lower bandwidth consumption, also without the need for constant connectivity). This promises to significantly enhance industrial applications but requires suited development tools to deal with the complexity of the edge technologies and context. We propose an agile Jupyter Python notebook as a simple, manageable tool to efficiently and effectively develop microcontroller-based intelligent imaging classification sensors. The notebook implements a methodology involving hyperparameter tuning and comparison of different shallow and deep learning models, with quantization. It exports TensorFlow Lite models, deployable on several microcontroller families, and optionally exploits the STM32Cube.AI Developer Cloud service, which allows benchmarking the developed models on a set of real-world tiny hardware target platforms. Assessment concerns various types of metrics, both for machine learning (e.g., accuracy) and embedded systems (e.g., memory footprint, latency, energy consumption). We have verified the support for development effectiveness and efficiency on four ultra-low resolution image-classification datasets, with different levels of input and task complexity. In all cases, the tool was able to build microcontroller-deployment-ready, beyond the state-of-the-art models, within one hour on Google Colab CPUs.

The focus of this work is to design a deeply quantized anomaly detector of oil leaks that may happen at the junction between the wind turbine high-speed shaft and the external bracket of the power generator. We propose a block-based binary shallow echo state network (BBS-ESN) architecture belonging to the reservoir computing (RC) category and, as we believe, it also extends the extreme learning machines (ELM) domain. Furthermore, BBS-ESN performs binary block-based online training using fixed and minimal computational complexity to achieve low power consumption and deployability on an off-the-shelf micro-controller (MCU). This has been achieved through binarization of the images and 1-bit quantization of the network weights and activations. 3D rendering has been used to generate a novel publicly available dataset of photo-realistic images similar to those potentially acquired by image sensors on the field while monitoring the junction, without and with oil leaks. Extensive experimentation has been conducted using a STM32H743ZI2 MCU running at 480 MHz and the results achieved show an accurate identification of anomalies, with a reduced computational cost per image and memory occupancy. Based on the obtained results, we conclude that BBS-ESN is feasible on off-the-shelf 32 bit MCUs. Moreover, the solution is also scalable in the number of image cameras to be deployed and to achieve accurate and fast oil leak detections from different viewpoints.