VAP3: IoT and the Urban Ecosystem
Wednesday, 17 June 2020, 12:15-14:30 CEST, Recommended re-viewing, https://www.youtube.com/playlist?list=PLjQu6nB1DfNCQgOI8O1oWPacsaPr5xBuT
Wednesday, 17 June 2020, 12:15-16:00 CEST, Non-Live interaction (Chat), link sent only to Registered people
Ignacio Martinez-Alpiste, Gelayol Golcarenarenji and Qi Wang (University of the West of Scotland, United Kingdom (Great Britain)); Jose Maria Alcaraz Calero (University of the West of Scotland & School of Engineering and Computing, United Kingdom (Great Britain))
Human Search and Rescue (SAR) tasks are mission-critical and take place in the wild, and thus solutions require timely and accurate human detection on a highly portable platform. This paper proposes a novel lightweight and practical SAR system that meets those demanding requirements by running optimised machine learning in a smartphone, interoperable with Unmanned Aerial Vehicles (UAV) that provides live video feed. In particular, the proposed approach significantly extends a standard machine learning algorithm to achieve adaptive object recognition in response to changing altitudes to accelerate the speed of finding missing people and eliminate redundant computing. Our approach achieved 91.02\% of accuracy and real-time speed on a smartphone that hosts the machine learning platform and the new algorithm. This proposed system is highly portable, cost-effective, fast with high accuracy suitable for UAV applications.
Muhammad Febrian Ardiansyah, Timothy William, Osamah Ibrahiem, Li-Chun Wang, Po-Lung Tien and Maria C. Yuang (National Chiao Tung University, Taiwan)
The fifth generation (5G) mobile network has paved the way for innovations across vertical industries. The integration of mobile edge computing to the design of the 5G end-to-end orchestrated architecture brings virtualization and low-latency computing to industries such as autonomous drone surveillance and navigation. In this paper, we present EagleEYE which stands for aerial edge-enabled disaster relief response system. First, EagleEYE employs a mechanism to combine two publicly available datasets to avoid creating an new dataset for emergency detection. Next, a deep learning object detection process is parallelized in the edge using containerization. Our experimental results show that EagleEYE can not only reduce the inference latency by 90% but also has high detection accuracy.
Kalkidan Gebru, Claudio E. Casetti, Carla Fabiana Chiasserini and Paolo Giaccone (Politecnico di Torino, Italy)
“The proliferation of IoT devices and the growing deployment of 5G networks combine to provide the perfect ecosystem for advanced smart city use cases. In this paper, we address the possibility of detecting and quantifying flows of people on city streets thanks to deployment of commercial sensors, connected to the 5G network, that capture WiFi probes transmitted by people’s smartphones. We first outline the motivation and challenges of such a scenario. Then, we illustrate our approach and present results derived from live measurements in a testbed deployed in the city of Turin within the 5G-EVE project. We show that we can quite accurately estimate transit flows by simply collecting anonymized MAC addresses and timestamps from smartphones of passers-by.”
Luis Diez (University of Cantabria, Spain); Ignacio Elicegui (Universidad de Cantabria, Spain); Luis Sanchez and Luis Muñoz (University of Cantabria, Spain)
Worldwide cities are involved in a digital transformation phase. More sustainable cities and improving citizen’s quality of life are the leit motiv of such transformation. However, such aims are difficult to achieve if the migration of the urban processes are not carried out following a common approach. Optimizing the behavior of any specific urban service needs to be performed taking into consideration both the service itself as well as its interaction with adjacent services. This means that any solution aiming to achieve the autonomous city management paradigm is tightly related to the adoption of common frameworks which are able to guarantee interoperability with other systems. Furthermore, cities themselves are not isolated systems. Well the opposite, cities interact one to the each other depending on different attributes. This implies that sooner or later optimizing some processes in one city without having in mind the adjacency to others might not be efficient enough. Hence, interoperability among cities will become a must, not just in terms of optimization but also replicability. Based on this boundary conditions this paper describes a framework aimed to ensure interoperability and replicability among cities. Some of the tools for assessing compliance with specific standardization activities are also presented.
Ahmed Abada and Marc St-Hilaire (Carleton University, Canada)
Since modern utility grids lack large-scale energy storage capabilities, their supply and demand levels must always be balanced to maintain a reliable operation. This requirement has traditionally been fulfilled through several existing techniques. However, recent interests in greening the utility grids require using more renewable energy sources, which makes the balancing task more challenging due to the unpredictability of such sources. This necessitates enhancing the grid’s flexibility (energy balancing capabilities) to ensure reliable grid operation. In this work, we propose a new energy auctioning mechanism that enhances the grid’s upward flexibility by using cloud datacenters as managed loads to quickly balance excess renewable energy. The auction process increases the energy consumption at a certain datacenter by incentivizing workload migrations to it in order to consume the excess energy. We present the formulation of our system model, introduce the used auction mechanism and show through simulation that it achieves efficient grid balancing, generates revenue on the sale of the excess energy and guarantees positive utility for all auction participants.