Article
Geosciences, Multidisciplinary
Beth Barnes, Sarah Dunn, Christopher Pearson, Sean Wilkinson
Summary: Disasters have significant social impacts, with emergency professionals needing proper assessment methods. This research aims to develop an agent-based modelling tool with robust human behavior models to assist in emergency contingency planning. By enhancing representation of human behavior, more accurate predictions of evacuation time can be achieved.
INTERNATIONAL JOURNAL OF DISASTER RISK REDUCTION
(2021)
Article
Operations Research & Management Science
Xiaojing Zheng
Summary: This paper introduces a new research method called exploratory computational experiment to solve problems from social complex systems. The method can describe social problems more accurately and provide more profound conclusions.
ANNALS OF OPERATIONS RESEARCH
(2023)
Article
Multidisciplinary Sciences
Kinga Makovi, Anahit Sargsyan, Wendi Li, Jean-Francois Bonnefon, Talal Rahwan
Summary: This study explores the differences in treatment between machines and humans in mixed collectives and the impact of following norms on trust. It suggests that humans are likely to rely on existing norms to develop cooperative norms in human-machine collectives.
NATURE COMMUNICATIONS
(2023)
Article
Management
Hyun-Rok Lee, Taesik Lee
Summary: This study addresses the issue of multiple decision-makers in disaster response using a decentralized-partially observable Markov decision process model. The proposed MARL algorithm augmented by pretraining neural network shows effectiveness and advantages in solving dec-POMDP problems.
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH
(2021)
Article
Geosciences, Multidisciplinary
Shengnan Wu, Yu Lei, Saini Yang, Peng Cui, Wen Jin
Summary: This study developed an agent-based model to integrate dynamic human behaviors into disaster risk management measures and evaluated their effectiveness in casualty reduction. The findings suggest that early warning systems are effective in community-based disaster risk management, with their credibility playing a critical role in their effectiveness. Additionally, early warning systems can be supplemented by other measures to further reduce casualties, although these measures may have negative effects on the effectiveness of the early warning system.
FRONTIERS IN EARTH SCIENCE
(2022)
Article
Computer Science, Hardware & Architecture
Sixia Fan, Qicai Zhou
Summary: New monitoring technology and instruments are needed to ensure the tunnel operates sustainably in economic, environmental, and engineering safety conditions. The system should provide comprehensive evaluation, timely warnings, and support decision-making with reduced costs and personnel.
Article
Thermodynamics
Ke Li, Ning Ye, Shuzhen Li, Haiyang Wang, Chenghui Zhang
Summary: As the conflict between energy and the environment intensifies, integrated energy system (IES) is an effective solution to address the difficulties of multi-energy coupling and multi-agent. This study proposes a multi-agent game operation strategy with integrated demand response (IDR) to optimize energy supply and utilization. The proposed distributed bi-level optimization model is solved using a distributed algorithm through genetic algorithm nested quadratic programming. Case studies in three scenarios demonstrate the effectiveness of the proposed method in improving supply-side revenue, reducing demand-side costs, and enhancing the operation and stability of energy supply and utilization.
Article
Environmental Sciences
Rohan T. Bhowmik, Youn Soo Jung, Juan A. Aguilera, Mary Prunicki, Kari Nadeau
Summary: Wildfires have a significant impact on the environment and human health, and California has experienced a surge in wildfires in recent years. This project developed a multi-modal wildfire prediction and early warning system using a novel spatio-temporal machine learning architecture. By integrating various data sources, including historical wildfires, environmental sensor data, and geological data, the system achieved a high accuracy rate in predicting and classifying wildfires. The system's implementation could save lives and protect the environment by providing early warnings and enabling better preparation.
JOURNAL OF ENVIRONMENTAL MANAGEMENT
(2023)
Article
Business
Linda Elmhadhbi, Mohamed-Hedi Karray, Bernard Archimede, J. Neil Otte, Barry Smith
Summary: Disaster response is a collaborative and critical process that involves multiple emergency responders. However, inadequate communication and interoperability issues can hinder the effectiveness of disaster response efforts. This article presents a scenario-based terrorism case study and proposes a semantics-based common operational command system to enhance information flow among emergency responders.
IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT
(2022)
Article
Thermodynamics
Jihong Ye, Wei Jiang, Xinxiang Yang, Bingyuan Hong
Summary: With the expansion of the petrochemical industry, safety production accidents have had a serious impact on people's lives and property. This paper proposes an emergency response framework for configuring and managing emergency supplies, using a multi-objective optimization method. The framework considers the pre-disaster and post-disaster stages and provides solutions for material configuration and delivery optimization. The study demonstrates the validity and practicality of the framework using a petrochemical enterprise in Zhoushan, Zhejiang as an example. It shows that the framework can reduce the safety and environmental impact of accidents and improve commodity scheduling efficiency.
Article
Computer Science, Artificial Intelligence
Li Chen, Yulong Zhang, Yanghe Feng, Longfei Zhang, Zhong Liu
Summary: In order to meet the high accuracy and low cost requirements of target classification in modern warfare and lay the foundation for target threat assessment, this article proposes a human-machine agent for target classification based on active reinforcement learning (TCARL_H-M), which can autonomously classify detected targets into predefined categories with equipment information and introduce human experience guidance for model. The article also presents a machine-based learner (TCARL_M) without human participation and a human-based interventionist (TCARL_H) with full human guidance to analyze the roles of human experience guidance and machine data learning in target classification tasks. The results from performance evaluation and application analysis demonstrate that TCARL_H-M not only greatly saves labor costs but also achieves more competitive classification accuracy compared to other models.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Information Systems
Minh-Ngoc Tran, Younghan Kim
Summary: Effective communication and information sharing are crucial in successful disaster response, with Named Data Networking (NDN) showing advantages over traditional IP networks. Real implementation of NDN in disaster scenarios is essential to ensure efficient communication and address challenges like responder mobility and network replacement.
Article
Multidisciplinary Sciences
Junqing Tang, Pengjun Zhao, Zhaoya Gong, Hongbo Zhao, Fengjue Huang, Jiaying Li, Zhihe Chen, Ling Yu, Jun Chen
Summary: Large-scale disasters disproportionately impact different population groups, especially the vulnerable and marginalized. This study investigates the resilience of human mobility during the unprecedented '720' Zhengzhou flood in China in 2021 using a massive amount of mobile phone signaling data. The findings reveal three counter-intuitive resilience patterns in human mobility and demonstrate that these patterns are not associated with gender or age.
NATIONAL SCIENCE REVIEW
(2023)
Article
Chemistry, Analytical
Amna Batool, Seng W. Loke, Niroshinie Fernando, Jonathan Kua
Summary: This paper proposes a methodology for making smart devices more human-centered by embedding ethical concepts and mapping them with socio-ethical policies. The methodology has been applied to various settings and devices, including robots, smart cameras, and smart speakers, resulting in significant improvement in their behavior towards human-centricity and adherence to ethical policies.
Article
Public, Environmental & Occupational Health
Ruben De Rouck, Mehdi Benhassine, Michel Debacker, Christian Dugauquier, Erwin Dhondt, Filip Van Utterbeeck, Ives Hubloue
Summary: In recent decades, there has been an increasing focus on preparedness for Chemical, Biological, Radiological, and Nuclear (CBRN) threats. Computer simulation is an effective tool for assessing contingency plans and improving response strategies. This paper presents a set of civilian nerve agent injury profiles based on military profiles, adapted for the civilian population. The methodology used can be applied to other chemical warfare agents and different exposure scenarios. These injury profiles can also be used in tabletop and live simulation exercises.
FRONTIERS IN PUBLIC HEALTH
(2023)
Article
Computer Science, Artificial Intelligence
Nhat Van-Quoc Truong, Le Cong Dinh, Sebastian Stein, Long Tran-Thanh, Nicholas R. Jennings
Summary: The success of crowdsourcing projects relies on motivating the crowd to contribute, which can be achieved through competitions and incentives. However, the best way to implement these contests in a particular project is still unknown. Therefore, a practical approach is to choose a set of incentives based on previous studies or the requester's experience, and use an effective mechanism to select appropriate incentives over time intervals to maximize overall utility.
APPLIED INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Jintong Ren, Jin-Kao Hao, Feng Wu, Zhang-Hua Fu
Summary: In this study, an intensification-driven local search algorithm is proposed to solve the Traveling Repairman Problem with Profits. By intensively investigating the areas around very-high-quality local optima, the algorithm obtains high-quality solutions. Experimental results show that the algorithm performs well by improving 36 best-known results and achieving equal best-known results for 95 instances out of 140 benchmark instances.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Cybernetics
Katie J. Parnell, Joel E. Fischer, Jediah R. Clark, Adrian Bodenmann, Maria Jose Galvez Trigo, Mario P. Brito, Mohammad Divband Soorati, Katherine L. Plant, Sarvapali D. Ramchurn
Summary: This study aims to understand the trust requirements of UAV operators when piloting UAVs by interviewing six operators with different levels of experience. The importance of past experience to trust and the expectations held by operators were identified, and recommendations were made for training and improving equipment, procedures and standards to develop trustworthy systems. The methodology developed shows promise for capturing trust in human-automation interactions.
INTERNATIONAL JOURNAL OF HUMAN-COMPUTER INTERACTION
(2023)
Article
Computer Science, Information Systems
Shreshth Tuli, Giuliano Casale, Nicholas R. Jennings
Summary: Edge Federation is a new computing paradigm that connects multiple edge service providers seamlessly. A challenge in such systems is deploying latency-critical and resource-intensive applications on constrained devices. To tackle this, a memory-efficient deep learning model called generative optimization networks (GON) is proposed. Leveraging the low memory footprint of GONs, a decentralized fault-tolerance method called DRAGON is introduced, which uses simulations to predict and optimize the performance of edge federation. Experimental results demonstrate that DRAGON outperforms baseline methods in fault-detection and Quality of Service (QoS) metrics, leading to improvements in energy consumption, response time, and service level agreement violations.
IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT
(2023)
Article
Oncology
Navamayooran Thavanesan, Indu Bodala, Zoe Walters, Sarvapali Ramchurn, Timothy J. Underwood, Ganesh Vigneswaran
Summary: This experimental pilot study developed machine learning models to predict treatment decisions in oesophageal cancer multidisciplinary teams. Results showed that multinomial logistic regression outperformed other algorithms in terms of performance metrics. Age was identified as a major factor in the decision-making process.
Review
Computer Science, Hardware & Architecture
Shreshth Tuli, Fatemeh Mirhakimi, Samodha Pallewatta, Syed Zawad, Giuliano Casale, Bahman Javadi, Feng Yan, Rajkumar Buyya, Nicholas R. Jennings
Summary: In recent years, there has been a shift in computing paradigms towards decentralized systems like IoT, Edge, Fog, Cloud, and Serverless. This shift has been powered by the adoption of AI-driven autonomous systems for managing distributed computing resources. This survey explores the evolution of data-driven AI methods and their impact on computing systems, focusing on resource management and QoS optimization. It also discusses future research directions and the potential of AI-driven computing systems.
JOURNAL OF NETWORK AND COMPUTER APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Farzaneh Farhadi, Maria Chli, Nicholas R. Jennings
Summary: This study explores an outsourcing problem where a software agent procures multiple services to complete a computational task before a deadline. The objective is to design an outsourcing strategy to maximize a specific objective function that depends on providers' costs. The study proposes a unified approach to create truthful procurement auctions for both socially-focused and self-interested consumers, using a weighted threshold payment scheme. The study also presents different procurement auctions that maximize expected utility and social welfare, assessed through game-theoretical and empirical analysis.
JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH
(2023)
Article
Computer Science, Hardware & Architecture
Shreshth Tuli, Giuliano Casale, Nicholas R. Jennings
Summary: The rise of distributed cloud computing technologies has been crucial for the widespread implementation of AI-based applications. However, existing AI solutions often fail to consider the interdependencies among different system performance aspects, leading to suboptimal performance in large cloud systems. To address this issue, a novel method called SciNet is proposed, which utilizes a co-simulated digital-twin to capture inter-metric dependencies and accurately estimate QoS scores. Experimental results demonstrate that SciNet outperforms state-of-the-art methods in terms of execution cost, inference accuracy, energy consumption, and response times.
IEEE TRANSACTIONS ON COMPUTERS
(2023)
Review
International Relations
Eryn Rigley, Caitlin Bentley, Joshua Krook, Sarvapali D. Ramchurn
Summary: As AI continues to disrupt industries, there is a skills gap in AI roles recruitment and employment for new graduates. The approaches to AI skills programs in different countries can guide future policy development for a skilled workforce and economic opportunities. The study found that countries emphasizing a broader, nationwide approach to educate all citizens had higher AI readiness scores compared to those focusing on a smaller group of AI experts. The findings suggest that future AI skills policy should adopt a broad, nationwide approach to upskill citizens at all levels.
Proceedings Paper
Computer Science, Artificial Intelligence
Donald McMillan, Razan Jaber, Benjamin R. Cowan, Joel E. Fischer, Bahar Irfan, Ronald Cumbal, Nima Zargham, Minha Lee
Summary: Conversation is a vital method of interaction between humans and robots, promoting inclusivity in Human-Robot Interaction (HRI). Both HRI and Conversational User Interfaces (CUI) research have contributed significantly to the design, understanding, and evaluation of human-robot conversational interactions. This workshop aims to bring together researchers from both fields to discuss shared opportunities and challenges in developing conversational interactions with robots and produce collaborative publications.
COMPANION OF THE ACM/IEEE INTERNATIONAL CONFERENCE ON HUMAN-ROBOT INTERACTION, HRI 2023
(2023)
Proceedings Paper
Automation & Control Systems
Gisela Reyes-Cruz, Isaac Phypers, Andriana Boudouraki, Dominic Price, Joel Fischer, Stuart Reeves, Maria Galvez Trigo, Horia Maior
Summary: Mobile robotic telepresence (MRP) provides remote users with access and mobility in various local environments. However, existing telepresence robots have limited physical manipulation capabilities, while Augmented Robotic Telepresence (ART) broadens the space for interaction and participation by augmenting the local environment through technology.
FIRST INTERNATIONAL SYMPOSIUM ON TRUSTWORTHY AUTONOMOUS SYSTEMS, TAS 2023
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Shreshth Tuli, Giuliano Casale, Nicholas R. Jennings
Summary: The paper proposes PreGAN, a composite AI model that uses a Generative Adversarial Network (GAN) for predicting preemptive migration decisions in containerized edge deployments for proactive fault-tolerance. Extensive experiments show that PreGAN outperforms existing methods in fault detection, diagnosis, and classification.
IEEE CONFERENCE ON COMPUTER COMMUNICATIONS (IEEE INFOCOM 2022)
(2022)
Proceedings Paper
Automation & Control Systems
Thomas G. Kelly, Mohammad Divband Soorati, Klaus-Peter Zauner, Sarvapali D. Ramchurn, Danesh Tarapore
Summary: This paper proposes a decentralized algorithm to improve collective decision making in communication-limited environments for robot swarms, without prior knowledge of the communication landscape. Our results show that making the swarm aware of the communication environment can significantly improve convergence speed, at least 3 times faster, without sacrificing accuracy.
2022 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Shreshth Tuli, Giuliano Casale, Nicholas R. Jennings
Summary: This study focuses on optimizing task scheduling in cloud computing environments and proposes a deep neural networks (DNNs) based method to improve execution costs. By using the surrogate model MetaNet, online dynamic selection of scheduling policy is achieved, resulting in optimization of task scheduling and execution costs.
2022 IEEE 15TH INTERNATIONAL CONFERENCE ON CLOUD COMPUTING (IEEE CLOUD 2022)
(2022)
Proceedings Paper
Education, Scientific Disciplines
Mohammad Naiseh, Caitlin Bentley, Sarvapali D. Ramehurn
Summary: Recent advances in artificial intelligence, especially machine learning, have positively contributed to the autonomous systems industry but have also brought about challenges in social, technical, legal, and ethical aspects. Trustworthy Autonomous Systems (TAS) is a researched and growing direction discussed in various disciplines, yet its impact on education curricula and required skills for future engineers remains rarely discussed. This study presents insights from TAS experts to highlight challenges in curriculum design and recommended skills for TAS education.
PROCEEDINGS OF THE 2022 IEEE GLOBAL ENGINEERING EDUCATION CONFERENCE (EDUCON 2022)
(2022)