Mobile users can open program in new tab for better viewing.

Open program in new tab

Day 1 02/12/2020
Day 2 03/12/2020
Day 3 04/12/2020
Room #1

Welcome Message by the General Chair 09:00 - 09:05

Prof. Sérgio Ivan Lopes, IT/IPVC

Welcome Message by EAI Conference Manager 09:05 - 09:10

Elena Davydova

Welcome Message by EAI Community Manager 09:10 - 09:15

Michal Dudic

Session 1: Edge-IoT: Computing & Edge AI 09:15 - 10:50

09:15 - 09:40
Technologies for Industrial Internet of Things (IIoT): Guidelines for Edge Computing Adoption in the Industry

The industrial sector is reinventing itself to find improved production processes. Industry 4.0 brings opportunities for growth productivity with the IIoT. There is a variety of IIoT-related technologies to manage vast volumes of data that should be stored and processed for the decision-making process. Cloud computing is a reliable option for remote storage and data processing. However, situations where the requirement is response time, intermittent connectivity, and low latency, edge computing processing is a more suitable option. Any device with computational resources can implement an edge computing capability with functions like a gateway and, and data aggregator capabilities. These devices need to be classified to avoid an uncontrolled proliferation of appliances in the enterprise's environment, which easily create cybersecurity vulnerabilities and transform device management into chaos. This study proposes a taxonomy for edge computing and defines rules for industrial adoption as a strategy to facilitate their sustainable implementation.
Authors: Ralf Moura (Vale S.A), Tiago Brasil (Vale S.A.), Ludmilla Werner (Vale S.A.), Alexandre Gonzalez (Vale S.A.), Claudio Dal' Cól (Vale S.A.), Sajjad Quadri (Vale S.A.),
Hide Authors & Abstract

Show Authors & Abstract
09:40 - 10:05
Scalable Approximate Computing Techniques for Latency and Bandwidth Constrained IoT Edge

Machine vision applications at the IoT Edge have bandwdith and latency constraints due to large sizes of video data. In this paper we propose approximate computing, that trades off inference accuracy with video frame size, as a potential solution. We present a number of low compute overhead video frame modifications that can reduce the video frame size, while achieving acceptable levels of inference accuracy. We present, a heuristic based design space pruning, and a Categorical boost based machine learning model as two approaches to achieve scalable performance in determining the appropriate video frame modifications that satisfy design constraints. Experimental results on an object detection application on the Microsoft COCO 2017 data set, indicates that proposed methods were able to reduce the video frame size by upto 71.3% while achieving an inference accuracy of 80.9% of that of the unmodified video frames. The machine learning model has a high training cost, but has a lower inference time, and is scalable and flexible compared to the heuristic design space pruning algorithm.
Authors: Anjus George (University of North Carolina at Charlotte), Arun Ravindran (University of North Carolina at Charlotte),
Hide Authors & Abstract

Show Authors & Abstract
10:05 - 10:30
Collaborative task processing with Internet of Things (IoT) Clusters

In this paper we propose a framework for resource sharing among Internet of Things (IoT) devices. Our goal is to optimize the use of this type of devices and specially those being underutilized. The infrastructure bellow our framework is typically built with heterogenous appliances that have specific functions and the idea is to minimize the need for software updates and other changes, trying to use the maximum of their spare resources with minimal interference. We base our solution in a pragmatic approach to task offloading based in the Erlang programming language.
Authors: Jorge Coelho (LIACC/UP ISEP/IPP), Luis Nogueira (ISEP/IPP),
Hide Authors & Abstract

Show Authors & Abstract
10:30 - 10:50
Inference Performance Comparison of Convolutional Neural Networks on Edge Devices

With the proliferation of Internet of Things (IoT), large amount of data are generated at edge devices with an unprecedented speed. In order to protect the privacy and security of big edge data, as well as reduce the communications cost, it is desirable to process the data locally at the edge devices. In this study, the inference performance of several popular pre-trained convolutional neural networks on three edge computing devices are evaluated. Specifically, MobileNetV1 & V2 and InceptionV3 models have been tested on NVIDIA Jetson TX2, Jetson Nano, and Google Edge TPU for image classification. Furthermore, various compression techniques including pruning, quantization, binarized neural network, and tensor decomposition are applied to reduce the model complexity. The results will provide a guidance for practitioners when deploying deep learning models on resource constrained edge devices for near real-time and on-site learning.
Authors: Sheikh Rufsan Reza (Prairie View A&M University), Yuzhong Yan (Prairie View A&M University), Xishuang Dong (Prairie View A&M University), Lijun Qian (Prairie View A&M University),
Hide Authors & Abstract

Show Authors & Abstract

Coffee Break 10:50 - 11:00

10 minutes

Session 2: Edge-IoT: Ecosystems & Applications 11:00 - 12:30

11:00 - 11:30
An Energy Sustainable CPS/IoT Ecosystem

This paper provides a short overview on methods and technologies necessary to build smart and sustainable Internet-of-Things (IoT). It observes IoT systems in a close relation with data centered intelligence and its application in cyber-physical systems. With the current rate of growth IoT devices and supporting CPS infrastructure will reach extremely high numbers in less than a decade. This will create an enormous overhead on world's supply of electrical energy. In this paper we propose a model extension for estimation of energy consumption by IoT devices in next decade. The paper gives a definition of CPS/IoT Ecosystem as a mutually codependent heterogeneous multidisciplinary structure. Further we explore a set of methods to reduce energy consumption and make CPS/IoT Ecosystem more sustainable. As a case study we propose energy harvesting sensor node and the corresponding use case in wildfire early detection systems.
Authors: Haris Isakovic (Technsiche Universität Wien), Edgar Crespo (Mondragon University), Radu Grosu (Technische Universität Wien),
Hide Authors & Abstract

Show Authors & Abstract
11:30 - 12:00
Low-cost LoRa-based IoT Edge Device for Indoor Air Quality Management in Schools

The air is an essential requirement for life, and its pollution creates serious problems for human health and well-being. Bearing in mind that the majority of society spends a large part of their time inside enclosed spaces, such as housing, educational establishments and other places of the same typology, Indoor Air Quality (IAQ) must be analyzed and taken into account, as a way to increase the level of quality of life of the subjects. Once the children, due to their characteristics, are more susceptible to developing health problems, it is essential to improve the school environment, namely, the air quality in classrooms. This work aims to describe the design and development of a LoRa-based IoT Edge device for classroom IAQ monitoring using low-cost commercial off-the-shelf components, capable of measuring relevant IAQ parameters specifically selected for the application case under study, namely carbon dioxide, particle matter, and volatile organic compounds. Lastly, the prototype is presented and evaluated in controlled conditions being its overall cost of approximately 150e.
Authors: António Abreu (Instituto Politécnico de Viana do Castelo), Sérgio Lopes (Instituto de Telecomunicações, Campus Universitário de Santiago, Aveiro, Portugal), Vítor Manso (DigiHeart - Consultadoria e Serviços em Tecnologias de Informação, Viana do Castelo, Portugal), António Curado (Instituto Politécnico de Viana do Castelo),
Hide Authors & Abstract

Show Authors & Abstract
12:00 - 12:30
NetButler: Voice-Based Edge/Cloud Virtual Assistant for Home Network Management

Virtual assistants (VA) are becoming a standard tool in many aspects of our daily lives that require technical support. Voice-based VAs in particular, such as Amazon Alexa and Google Assistant, have become common in smart phones and domestic IoT devices (e.g., Google Home, Amazon Echo), replying to user inquiries (e.g., weather forecast) or performing simple services (e.g., play music). Through dedicated interfaces, VAs can be used or extended to support new services, and one particular area typically requiring assistance is the management of home networks. Activating specific features or troubleshooting connectivity problems may be difficult or impossible for users that are not tech-savvy. In this paper we introduce NetButler, a voice-based virtual assistant tailored to support the management of home networks, that leverages a third-party cloud-based voice service (Alexa) and dedicated routines at the home gateway. Offered functionalities are the setup of a guest network and diagnosis of connectivity problems, by quantifying the signal strength of the devices in the local network and performing a throughput test to an external server. We evaluate the user experience with the NetButler system with 8 test users. We report an average of up to 15s to set up a guest network and between 30 to 60s to diagnose various problems, and we find overall user satisfaction to be 3.75 in a 1-to-5 scale by means of a after-interaction questionnaire.
Authors: Diogo Martins (Universidade do Porto, Faculdade de Engenharia (FEUP)), Bruno Parreira (Altran S.A.), Pedro Miguel Santos (CISTER research center, Instituto Politécnico do Porto (CISTER/IPP); Faculdade de Engenharia, Universidade do Porto (FE-UP)), Sérgio Figueiredo (Altran S.A.),
Hide Authors & Abstract

Show Authors & Abstract

Lunch 12:30 - 13:00

30 minutes

Session 2: Edge-IoT: Ecosystems & Applications 13:00 - 13:25

13:00 - 13:25
Environment Monitoring Modules with Fire Detection Capability based on IoT Methodology

Worldwide, forests have been devastated by fires in recent years. Whe- ther by human intervention or for other reasons, the history of burned areas is increasing year after year, degrading fauna and flora. For this reason, it is vital to detect an early ignition so that firefighters can act quickly, reducing the impacts caused by forest fires. The proposed system aims to improve the nature monitoring and to assist the existing surveillance systems through Wireless Sensor Network. The network formed by the set of sensors has the potential to identify forest ignitions and, consequently, alerts the authorities through LoRaWAN communication. This work presents a prototype based on low-cost technology, which can be used in areas that require a high density of modules. Tests with a Wireless Sensor Network made up of nine prototypes demonstrate its effectiveness and robustness in terms of data transmission and collection. In this way, it is possible to apply this approach in Portuguese forests with a high level of forest fire risk, transforming them into Forests 4.0 concept.
Authors: Thadeu Brito (, Beatriz Azevedo (CeDRI), Antonio Valente (Universidade de Trás-os-Montes e Alto Douro, Portugal), Ana Pereira (CeDRI), Jose Lima (CeDRI), Paulo Costa (FEUP),
Hide Authors & Abstract

Show Authors & Abstract

Keynote: Assoc. Prof. António Valente 13:25 - 13:55

Enabling the Douro Demarcated Region for the Internet of Things

Conference Wrap Up 13:55 - 14:00

Prof. Sérgio Ivan Lopes, IT/IPVC