Environmental Sound Recognition (ESR) plays a crucial role in smart cities by accurately categorizing audio using well-trained Machine Learning (ML) classifiers. This application is particularly valuable for cities that analyzed environmental sounds to gain insight and data. However, deploying deep learning (DL) models on resource-constrained embedded devices, such as Raspberry Pi (RPi) or Tensor Processing Units (TPUs), poses challenges. In this work, an evaluation of an existing pre-trained model for deployment on Raspberry Pi (RPi) and TPU platforms other than a laptop is proposed. We explored the impact of the retraining parameters and compared the sound classification performance across three datasets: ESC-10, BDLib, and Urban Sound. Our results demonstrate the effectiveness of the pre-trained model for transfer learning in embedded systems. On laptops, the accuracy rates reached 96.6% for ESC-10, 100% for BDLib, and 99% for Urban Sound. On RPi, the accuracy rates were 96.4% for ESC-10, 100% for BDLib, and 95.3% for Urban Sound, while on RPi with Coral TPU, the rates were 95.7% for ESC-10, 100% for BDLib and 95.4% for the Urban Sound. Utilizing pre-trained models reduces the computational requirements, enabling faster inference. Leveraging pre-trained models in embedded systems accelerates the development, deployment, and performance of various real-time applications.

In recent years, Environmental Sound Recognition (ESR) has become a relevant capability for urban monitoring applications. The techniques for automated sound recognition often rely on machine learning approaches, which have increased in complexity in order to achieve higher accuracy. Nonetheless, such machine learning techniques often have to be deployed on resource and power-constrained embedded devices, which has become a challenge with the adoption of deep learning approaches based on Convolutional Neural Networks (CNNs). Field-Programmable Gate Arrays (FPGAs) are power efficient and highly suitable for computationally intensive algorithms like CNNs. By fully exploiting their parallel nature, they have the potential to accelerate the inference time as compared to other embedded devices. Similarly, dedicated architectures to accelerate Artificial Intelligence (AI) such as Tensor Processing Units (TPUs) promise to deliver high accuracy while achieving high performance. In this work, we evaluate existing tool flows to deploy CNN models on FPGAs as well as on TPU platforms. We propose and adjust several CNN-based sound classifiers to be embedded on such hardware accelerators. The results demonstrate the maturity of the existing tools and how FPGAs can be exploited to outperform TPUs.

Water is a major preoccupation for our generation since it is crucial in keeping a healthy ecosystem and supporting biodiversity. The state of aquatic systems and water bodies needs to be continiously monitored to make informed decisions and trigger sanitation when necessary. However, observing and tracking the evolution of many water bodies without disturbing and polluting the biotopes is expensive, not scalable, and thus infeasible. This paper presents a way to make sustainable measurements using a new low-cost, open-source, and autonomous monitoring system deployable in a broad network. The smart buoy is deployed and controlled by a central unit that uses lab-graded sensors to measure ambient factors. The custom electronic board offers sustainable electronics integration emphasizing power path and network connectivity. The smart buoy showed an average power consumption of 1.8mA and a cost of 932 euros per device. Currently, five spots have been monitored, which allowed the understanding of why biological events, such as a massive fish dead, occurred. The system is easily expandable and can be used in various applications to increase the knowledge of the underwater ecosystem.

Environmental Sound Recognition has become a relevant application for smart cities. Such an application, however, demands the use of trained machine learning classifiers in order to categorize a limited set of audio categories. Although classical machine learning solutions have been proposed in the past, most of the latest solutions that have been proposed toward automated and accurate sound classification are based on a deep learning approach. Deep learning models tend to be large, which can be problematic when considering that sound classifiers often have to be embedded in resource constrained devices. In this paper, a classical machine learning based classifier called MosAIc, and a lighter Convolutional Neural Network model for environmental sound recognition, are proposed to directly compete in terms of accuracy with the latest deep learning solutions. Both approaches are evaluated in an embedded system in order to identify the key parameters when placing such applications on constrained devices. The experimental results show that classical machine learning classifiers can be combined to achieve similar results to deep learning models, and even outperform them in accuracy. The cost, however, is a larger classification time.

In recent years, smart water sensing technology has played a crucial role in water management, addressing the pressing need for efficient monitoring and control of water resources analysis. The challenge in smart water sensing technology resides in ensuring the reliability and accuracy of the data collected by sensors. Outliers are a well-known problem in smart sensing as they can negatively affect the viability of useful analysis and make it difficult to evaluate pertinent data. In this study, we evaluate the performance of four sensors: electrical conductivity (EC), dissolved oxygen (DO), temperature (Temp), and pH. We implement four classical machine learning models: support vector machine (SVM), artifical neural network (ANN), decision tree (DT), and isolated forest (iForest)-based outlier detection as a pre-processing step before visualizing the data. The dataset was collected by a real-time smart water sensing monitoring system installed in Brussels’ lakes, rivers, and ponds. The obtained results clearly show that the SVM outperforms the other models, showing 98.38% $F_1$-score rates for pH, 96.98% $F_1$-score rates for temp, 97.88% $F_1$-score rates for DO, and 98.11% $F_1$-score rates for EC. Furthermore, ANN also achieves a significant results, establishing it as a viable alternative.