Artificial Intelligence on the Edge

A special issue of Information (ISSN 2078-2489). This special issue belongs to the section "Artificial Intelligence".

Deadline for manuscript submissions: 31 July 2024 | Viewed by 24146

Special Issue Editors


E-Mail Website
Guest Editor
Department of Mathematics, Computer Science, Physics and Hearth Sciences (MIFT), University of Messina, 98166 Messina, Italy
Interests: artificial intelligence; cloud computing; edge computing; big data
Special Issues, Collections and Topics in MDPI journals

E-Mail Website
Guest Editor
MIFT Department, Universita degli Studi di Messina, Viale F. Stagno D'Alcontres 31, 98166 Messina, Italy
Interests: osmotic computing; cloud computing; Fog-Edge computing; IoT; IT security
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

As the demand for Internet of Things (IoT) solutions has grown over time, edge computing has gained momentum, because of the need to move computation close to data sources. Indeed, the rise of connected devices, which Gartner estimates number more than 20 quadrillion in 2020, is growing the need for computation in scenarios where prompt responses are crucial. On the other hand, the upcoming deployment of 5G networks will bring drastic performance improvements, traffic optimization and new ultra-low-latency services in locations where cloud connectivity is too low, such as in oil platforms or cruise ships.

On the other hand, the increasing number of connected devices is also generating a huge number of raw data directly on the edge. For example, Cisco estimates that nearly 850 ZB will be generated by people, machines and things at the network edge by 2021 [1]. That is where artificial intelligence (AI) steps in, leading data transformation in real-time extracted business value. Therefore, the migration of machine learning and deep learning techniques over to the edge enables a new field of research, where the intelligence is distributed over devices. In this context, TensorFlow has already released a tool that enables AI on the edge, but many challenges remain.

The benefits of AI on the edge are typically visible over several application fields, such as wearable technologies, smart homes, smart cities, Industry 4.0, agriculture, autonomous driving, video surveillance, social and industrial robotics, etc.

This Special Issue aims to promote high-quality research on all the aspects related to the training, inference and migration to the edge of artificial intelligence services. Topics of interest include, but are not limited to:

  • Machine learning services on the edge;
  • Deep learning services on the edge;
  • The migration of AI-based services from the cloud into the edge;
  • The optimization of real-time, AI-based solutions on the edge;
  • Edge-centric distributed intelligent services;
  • Edge-centric collaborative intelligent services;
  • Edge-centric federated intelligent services;
  • The security of data distribution over AI-based edge systems;
  • Trust and privacy management in AI-based edge systems;
  • The quality of services and energy efficiency for AI-based edge systems;
  • AI for the IoT;
  • AI for microcontroller and microprocessor.

Dr. Lorenzo Carnevale
Dr. Massimo Villari
Guest Editors

References:

  1. https://www.cisco.com/c/en/us/solutions/executive-perspectives/annual-internet-report/index.html

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Information is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Published Papers (9 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

Jump to: Review

19 pages, 2394 KiB  
Article
Reducing the Power Consumption of Edge Devices Supporting Ambient Intelligence Applications
by Anastasios Fanariotis, Theofanis Orphanoudakis and Vassilis Fotopoulos
Information 2024, 15(3), 161; https://doi.org/10.3390/info15030161 - 12 Mar 2024
Viewed by 1442
Abstract
Having as a main objective the exploration of power efficiency of microcontrollers running machine learning models, this manuscript contrasts the performance of two types of state-of-the-art microcontrollers, namely ESP32 with an LX6 core and ESP32-S3 with an LX7 core, focusing on the impact [...] Read more.
Having as a main objective the exploration of power efficiency of microcontrollers running machine learning models, this manuscript contrasts the performance of two types of state-of-the-art microcontrollers, namely ESP32 with an LX6 core and ESP32-S3 with an LX7 core, focusing on the impact of process acceleration technologies like cache memory and vectoring. The research employs experimental methods, where identical machine learning models are run on both microcontrollers under varying conditions, with particular attention to cache optimization and vector instruction utilization. Results indicate a notable difference in power efficiency between the two microcontrollers, directly linked to their respective process acceleration capabilities. The study concludes that while both microcontrollers show efficacy in running machine learning models, ESP32-S3 with an LX7 core demonstrates superior power efficiency, attributable to its advanced vector instruction set and optimized cache memory usage. These findings provide valuable insights for the design of power-efficient embedded systems supporting machine learning for a variety of applications, including IoT and wearable devices, ambient intelligence, and edge computing and pave the way for future research in optimizing machine learning models for low-power, embedded environments. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

18 pages, 1219 KiB  
Article
Neural Network-Based Solar Irradiance Forecast for Edge Computing Devices
by Georgios Venitourakis, Christoforos Vasilakis, Alexandros Tsagkaropoulos, Tzouma Amrou, Georgios Konstantoulakis, Panagiotis Golemis and Dionysios Reisis
Information 2023, 14(11), 617; https://doi.org/10.3390/info14110617 - 18 Nov 2023
Cited by 1 | Viewed by 1261
Abstract
Aiming at effectively improving photovoltaic (PV) park operation and the stability of the electricity grid, the current paper addresses the design and development of a novel system achieving the short-term irradiance forecasting for the PV park area, which is the key factor for [...] Read more.
Aiming at effectively improving photovoltaic (PV) park operation and the stability of the electricity grid, the current paper addresses the design and development of a novel system achieving the short-term irradiance forecasting for the PV park area, which is the key factor for controlling the variations in the PV power production. First, it introduces the Xception long short-term memory (XceptionLSTM) cell tailored for recurrent neural networks (RNN). Second, it presents the novel irradiance forecasting model that consists of a sequence-to-sequence image regression NNs in the form of a spatio-temporal encoder–decoder including Xception layers in the spatial encoder, the novel XceptionLSTM in the temporal encoder and decoder and a multilayer perceptron in the spatial decoder. The proposed model achieves a forecast skill of 16.57% for a horizon of 5 min when compared to the persistence model. Moreover, the proposed model is designed for execution on edge computing devices and the real-time application of the inference on the Raspberry Pi 4 Model B 8 GB and the Raspberry Pi Zero 2W validates the results. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

21 pages, 3951 KiB  
Article
Image-Based Insect Counting Embedded in E-Traps That Learn without Manual Image Annotation and Self-Dispose Captured Insects
by Ioannis Saradopoulos, Ilyas Potamitis, Antonios I. Konstantaras, Panagiotis Eliopoulos, Stavros Ntalampiras and Iraklis Rigakis
Information 2023, 14(5), 267; https://doi.org/10.3390/info14050267 - 30 Apr 2023
Cited by 1 | Viewed by 2850
Abstract
This study describes the development of an image-based insect trap diverging from the plug-in camera insect trap paradigm in that (a) it does not require manual annotation of images to learn how to count targeted pests, and (b) it self-disposes the captured insects, [...] Read more.
This study describes the development of an image-based insect trap diverging from the plug-in camera insect trap paradigm in that (a) it does not require manual annotation of images to learn how to count targeted pests, and (b) it self-disposes the captured insects, and therefore is suitable for long-term deployment. The device consists of an imaging sensor integrated with Raspberry Pi microcontroller units with embedded deep learning algorithms that count agricultural pests inside a pheromone-based funnel trap. The device also receives commands from the server, which configures its operation, while an embedded servomotor can automatically rotate the detached bottom of the bucket to dispose of dehydrated insects as they begin to pile up. Therefore, it completely overcomes a major limitation of camera-based insect traps: the inevitable overlap and occlusion caused by the decay and layering of insects during long-term operation, thus extending the autonomous operational capability. We study cases that are underrepresented in the literature such as counting in situations of congestion and significant debris using crowd counting algorithms encountered in human surveillance. Finally, we perform comparative analysis of the results from different deep learning approaches (YOLOv7/8, crowd counting, deep learning regression). Interestingly, there is no one optimal clear-cut counting approach that can cover all situations involving small and large insects with overlap. By weighting the pros and cons we suggest that YOLOv7/8 provides the best embedded solution in general. We open-source the code and a large database of Lepidopteran plant pests. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

13 pages, 2592 KiB  
Article
Efficient Dynamic Reconfigurable CNN Accelerator for Edge Intelligence Computing on FPGA
by Kaisheng Shi, Mingwei Wang, Xin Tan, Qianghua Li and Tao Lei
Information 2023, 14(3), 194; https://doi.org/10.3390/info14030194 - 20 Mar 2023
Cited by 3 | Viewed by 2453
Abstract
This paper proposes an efficient dynamic reconfigurable CNN accelerator (EDRCA) for FPGAs to tackle the issues of limited hardware resources and low energy efficiency in the deployment of convolutional neural networks on embedded edge computing devices. First, a configuration layer sequence optimization method [...] Read more.
This paper proposes an efficient dynamic reconfigurable CNN accelerator (EDRCA) for FPGAs to tackle the issues of limited hardware resources and low energy efficiency in the deployment of convolutional neural networks on embedded edge computing devices. First, a configuration layer sequence optimization method is proposed to minimize the configuration time overhead and improve performance. Second, accelerator templates for dynamic regions are designed to create a unified high-speed interface and enhance operational performance. The dynamic reconfigurable technology is applied on the Xilinx KV260 FPGA platform to design the EDRCA accelerator, resolving the hardware resource constraints in traditional accelerator design. The YOLOV2-TINY object detection network is used to test the EDRCA accelerator on the Xilinx KV260 platform using floating point data. Results at 250 MHz show a computing performance of 75.1929 GOPS, peak power consumption of 5.25 W, and power efficiency of 13.6219 GOPS/W, indicating the potential of the EDRCA accelerator for edge intelligence computing. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

15 pages, 1897 KiB  
Article
Earthquake Detection at the Edge: IoT Crowdsensing Network
by Enrico Bassetti and Emanuele Panizzi
Information 2022, 13(4), 195; https://doi.org/10.3390/info13040195 - 13 Apr 2022
Cited by 10 | Viewed by 3272
Abstract
State-of-the-art Earthquake Early Warning systems rely on a network of sensors connected to a fusion center in a client–server paradigm. The fusion center runs different algorithms on the whole data set to detect earthquakes. Instead, we propose moving computation to the edge, with [...] Read more.
State-of-the-art Earthquake Early Warning systems rely on a network of sensors connected to a fusion center in a client–server paradigm. The fusion center runs different algorithms on the whole data set to detect earthquakes. Instead, we propose moving computation to the edge, with detector nodes that probe the environment and process information from nearby probes to detect earthquakes locally. Our approach tolerates multiple node faults and partial network disruption and keeps all data locally, enhancing privacy. This paper describes our proposal’s rationale and explains its architecture. We then present an implementation that uses Raspberry, NodeMCU, and the Crowdquake machine learning model. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

16 pages, 3750 KiB  
Article
Shrink and Eliminate: A Study of Post-Training Quantization and Repeated Operations Elimination in RNN Models
by Nesma M. Rezk, Tomas Nordström and Zain Ul-Abdin
Information 2022, 13(4), 176; https://doi.org/10.3390/info13040176 - 31 Mar 2022
Cited by 2 | Viewed by 2164
Abstract
Recurrent neural networks (RNNs) are neural networks (NN) designed for time-series applications. There is a growing interest in running RNNs to support these applications on edge devices. However, RNNs have large memory and computational demands that make them challenging to implement on edge [...] Read more.
Recurrent neural networks (RNNs) are neural networks (NN) designed for time-series applications. There is a growing interest in running RNNs to support these applications on edge devices. However, RNNs have large memory and computational demands that make them challenging to implement on edge devices. Quantization is used to shrink the size and the computational needs of such models by decreasing weights and activation precision. Further, the delta networks method increases the sparsity in activation vectors by relying on the temporal relationship between successive input sequences to eliminate repeated computations and memory accesses. In this paper, we study the effect of quantization on LSTM-, GRU-, LiGRU-, and SRU-based RNN models for speech recognition on the TIMIT dataset. We show how to apply post-training quantization on these models with a minimal increase in the error by skipping quantization of selected paths. In addition, we show that the quantization of activation vectors in RNNs to integer precision leads to considerable sparsity if the delta networks method is applied. Then, we propose a method for increasing the sparsity in the activation vectors while minimizing the error and maximizing the percentage of eliminated computations. The proposed quantization method managed to compress the four models more than 85%, with an error increase of 0.6, 0, 2.1, and 0.2 percentage points, respectively. By applying the delta networks method to the quantized models, more than 50% of the operations can be eliminated, in most cases with only a minor increase in the error. Comparing the four models to each other under the quantization and delta networks method, we found that compressed LSTM-based models are the most-optimum solutions at low-error-rates constraints. The compressed SRU-based models are the smallest in size, suitable when higher error rates are acceptable, and the compressed LiGRU-based models have the highest number of eliminated operations. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

19 pages, 1440 KiB  
Article
File System Support for Privacy-Preserving Analysis and Forensics in Low-Bandwidth Edge Environments
by Aril Bernhard Ovesen, Tor-Arne Schmidt Nordmo, Håvard Dagenborg Johansen, Michael Alexander Riegler, Pål Halvorsen and Dag Johansen
Information 2021, 12(10), 430; https://doi.org/10.3390/info12100430 - 18 Oct 2021
Cited by 6 | Viewed by 2304
Abstract
In this paper, we present initial results from our distributed edge systems research in the domain of sustainable harvesting of common good resources in the Arctic Ocean. Specifically, we are developing a digital platform for real-time privacy-preserving sustainability management in the domain of [...] Read more.
In this paper, we present initial results from our distributed edge systems research in the domain of sustainable harvesting of common good resources in the Arctic Ocean. Specifically, we are developing a digital platform for real-time privacy-preserving sustainability management in the domain of commercial fishery surveillance operations. This is in response to potentially privacy-infringing mandates from some governments to combat overfishing and other sustainability challenges. Our approach is to deploy sensory devices and distributed artificial intelligence algorithms on mobile, offshore fishing vessels and at mainland central control centers. To facilitate this, we need a novel data plane supporting efficient, available, secure, tamper-proof, and compliant data management in this weakly connected offshore environment. We have built our first prototype of Dorvu, a novel distributed file system in this context. Our devised architecture, the design trade-offs among conflicting properties, and our initial experiences are further detailed in this paper. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

18 pages, 4166 KiB  
Article
Combine-Net: An Improved Filter Pruning Algorithm
by Jinghan Wang, Guangyue Li and Wenzhao Zhang
Information 2021, 12(7), 264; https://doi.org/10.3390/info12070264 - 29 Jun 2021
Cited by 2 | Viewed by 2865
Abstract
The powerful performance of deep learning is evident to all. With the deepening of research, neural networks have become more complex and not easily generalized to resource-constrained devices. The emergence of a series of model compression algorithms makes artificial intelligence on edge possible. [...] Read more.
The powerful performance of deep learning is evident to all. With the deepening of research, neural networks have become more complex and not easily generalized to resource-constrained devices. The emergence of a series of model compression algorithms makes artificial intelligence on edge possible. Among them, structured model pruning is widely utilized because of its versatility. Structured pruning prunes the neural network itself and discards some relatively unimportant structures to compress the model’s size. However, in the previous pruning work, problems such as evaluation errors of networks, empirical determination of pruning rate, and low retraining efficiency remain. Therefore, we propose an accurate, objective, and efficient pruning algorithm—Combine-Net, introducing Adaptive BN to eliminate evaluation errors, the Kneedle algorithm to determine the pruning rate objectively, and knowledge distillation to improve the efficiency of retraining. Results show that, without precision loss, Combine-Net achieves 95% parameter compression and 83% computation compression on VGG16 on CIFAR10, 71% of parameter compression and 41% computation compression on ResNet50 on CIFAR100. Experiments on different datasets and models have proved that Combine-Net can efficiently compress the neural network’s parameters and computation. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

Review

Jump to: Research

19 pages, 983 KiB  
Review
Advancements in On-Device Deep Neural Networks
by Kavya Saravanan and Abbas Z. Kouzani
Information 2023, 14(8), 470; https://doi.org/10.3390/info14080470 - 21 Aug 2023
Cited by 2 | Viewed by 1737
Abstract
In recent years, rapid advancements in both hardware and software technologies have resulted in the ability to execute artificial intelligence (AI) algorithms on low-resource devices. The combination of high-speed, low-power electronic hardware and efficient AI algorithms is driving the emergence of on-device AI. [...] Read more.
In recent years, rapid advancements in both hardware and software technologies have resulted in the ability to execute artificial intelligence (AI) algorithms on low-resource devices. The combination of high-speed, low-power electronic hardware and efficient AI algorithms is driving the emergence of on-device AI. Deep neural networks (DNNs) are highly effective AI algorithms used for identifying patterns in complex data. DNNs, however, contain many parameters and operations that make them computationally intensive to execute. Accordingly, DNNs are usually executed on high-resource backend processors. This causes an increase in data processing latency and energy expenditure. Therefore, modern strategies are being developed to facilitate the implementation of DNNs on devices with limited resources. This paper presents a detailed review of the current methods and structures that have been developed to deploy DNNs on devices with limited resources. Firstly, an overview of DNNs is presented. Next, the methods used to implement DNNs on resource-constrained devices are explained. Following this, the existing works reported in the literature on the execution of DNNs on low-resource devices are reviewed. The reviewed works are classified into three categories: software, hardware, and hardware/software co-design. Then, a discussion on the reviewed approaches is given, followed by a list of challenges and future prospects of on-device AI, together with its emerging applications. Full article
(This article belongs to the Special Issue Artificial Intelligence on the Edge)
Show Figures

Figure 1

Back to TopTop