Collaborative Intelligence Research Articles

A DNN partitioning framework with controlled lossy mechanisms for edge-cloud collaborative intelligence

Collaborative intelligence (CI) is an approach used to accelerate deep neural network computations on IoT edge devices by offloading certain DNN computations to high-performance cloud servers. This offloading process involves transferring a large volume of feature data between the IoT edge device and the cloud servers, resulting in significant transfer time overhead that contributes to the overall end-to-end latency. Despite its importance, CI research has largely overlooked the impact of an unstable network environment on transfer time, leading to suboptimal latency in real-world scenarios. In light of this, we propose CO-PILOT, a novel partition framework that aims to minimize end-to-end latency while preserving the original model accuracy. CO-PILOT achieves this by finely controlling various lossy mechanisms, such as compression level and transport protocol, and strategically selecting error-robust layers as partition points to enhance resilience against network conditions and lossy mechanisms. The results demonstrate that CO-PILOT can reduce end-to-end latency by up to 85% with less than 1% accuracy degradation, all without requiring model retraining.

Expert or machine? Comparing the effect of pairing student teacher with in-service teacher and ChatGPT on their critical thinking, learning performance, and cognitive load in an integrated-STEM course

ABSTRACT For student teachers’ professional development, the emergence of generative artificial intelligence (AI) represents both opportunity and challenge. This exploratory quasi-experimental study aims to investigate the effects of “Human-Human” and “Human-Machine” collaborative learning approaches on the SETM teaching training performance of student teachers. Twenty-three student teachers were divided into two groups within a single class, each adopting one learning method. The experiment lasted for two months with weekly three-hour sessions. Data were analysed focusing on critical thinking, learning performance, and cognitive load between the groups. The results indicated that student teachers using ChatGPT showed higher critical thinking systematicity, task completion efficiency, and experienced lower cognitive load. Student teachers paired with in-service teachers slightly outperformed those with ChatGPT on the final teaching design proposal. These findings underscore the potential and varying strengths of AI tools like ChatGPT and human teachers. For further research, refined collaborative learning scaffolding are recommended to explore the impact and potential of AI-assisted and in-service teacher-involved collaboration. The study’s implications could guide educators, policymakers, and AI developers in optimizing the AI-enhanced collaborative learning strategies and shed light on the new formation of human-machine collaborative intelligence in the scope of education.

Intelligent Collaboration and Control of Robots in Assembly Production Lines

Intelligent Collaboration and Control of Robots in Assembly Production Lines

Low-Rate Feature Compression for Collaborative Intelligence: Reducing Redundancy in Spatial and Statistical Levels

Low-Rate Feature Compression for Collaborative Intelligence: Reducing Redundancy in Spatial and Statistical Levels

Learning Human Behavior in Shared Control: Adaptive Inverse Differential Game Approach.

To enhance the collaborative intelligence of a machine, it is important for the machine to understand what behavior a human may adopt to interact with the machine when performing a task in shared control. In this study, an online behavior learning method is proposed for continuous-time linear human-in-the-loop shared control systems by using the system state data only. A two-player nonzero-sum linear quadratic dynamic game paradigm is used for modeling the control interaction between a human operator and an automation that actively compensates for human control action. In this game model, the cost function representing the human behavior is assumed to have an unknown weighting matrix. Here, we want to learn the human behavior or retrieve the weighting matrix by using the system state data only. Accordingly, a new adaptive inverse differential game (IDG) method, which integrates concurrent learning (CL) and linear matrix inequality (LMI) optimization, is proposed. First, a CL-based adaptive law and an interactive controller of the automation are developed to estimate the feedback gain matrix of the human online, and second, an LMI optimization problem is solved to determine the weighting matrix of the human cost function. Finally, simulation results on a cooperative shared control driver assistance system are provided to elucidate the feasibility of the developed method.

Explainable Automated Machine Learning for Credit Decisions: Enhancing Human Artificial Intelligence Collaboration in Financial Engineering

Explainable Automated Machine Learning for Credit Decisions: Enhancing Human Artificial Intelligence Collaboration in Financial Engineering

Study on the Regulation of Criminal Procedure System in the Age of Artificial Intelligence

Abstract The development of artificial intelligence technology has promoted economic development and improved people’s living standards, but it will also cause many risks and uncertainties. The purpose of this paper is to create a reciprocal fairness model for artificial intelligence collaboration and explore the causal relationship and liability allocation in artificial intelligence criminal infringement cases based on Rabin’s theory. Through the Hande formula, the criteria for determining fault in the case, as well as the marginal costs and benefits of the aggressor’s use of AI to commit crimes, are calculated. Finally, based on the theory of behavioral game evolution, this paper discusses the cooperation law of the three parties under the regulation of the Criminal Procedure Law, and discusses the illegal cost of the illegal implementation of artificial intelligence from the amount of compensation. The results show that in 68,535 cases of using AI to commit crimes in 2022, the average compensation is 74,556.87 yuan, indicating that the cost of crime is much lower than the proceeds of crime. The proposed legal regulations and the prevention and control of AI technology risks can be practiced through this study’s practical relevance.

Intelligence Collaboration as a Potential Mitigation of the Golden Visa Threat

The Golden Visa policy, aimed at attracting quality investors and global talents, has gained prominence globally, including in Indonesia through Permenkumham 22 of 2023. This policy is designed to enhance the domestic economy by offering incentives and streamlined processes for foreign investors and talents. However, the potential threat posed by the Golden Visa system is a subject of concern, prompting this research to delve into its implications using qualitative research methods and an inductive approach. The analysis reveals that the potential threat of the Golden Visa lies in its capacity to negatively impact national security. Recognizing the significant role of the Directorate General of Immigration, the agency vested with authority in the immigration sector, becomes crucial in mitigating these threats. The integration of intelligence analysis and collaborative governance within the Directorate General of Immigration can serve as an effective early detection mechanism for potential threats associated with the Golden Visa program. In conclusion, while the Golden Visa policy holds the promise of economic benefits for Indonesia, it is imperative to address and manage the potential security threats it may pose. By leveraging the expertise of immigration authorities and employing strategic intelligence analysis, Indonesia can strike a balance between attracting foreign investments and talents while safeguarding its national security interests.

An evaluation of factors influencing the community emergency management under compounding risks perspective

Communities in China are proactive in emergency response, significantly mitigating compounding risks and enhancing resilience. There exists a crucial need to analyze the factors influencing community emergency management, to strengthen the overall competencies of emergency response. This study examined the complex interrelationships among 15 identified factors using the Decision Making Trial and Evaluation Laboratory (DEMATEL) - Adversarial Interpretive Structure Model (AISM) modeling techniques. According to the hierarchical structure determined by the AISM, three main elements emerged as critical: emergency management systems, emergency governance models, and emergency intelligence collaboration systems. These findings offer valuable insights for community administrators, providing an in-depth understanding of the vulnerabilities in community emergency management. Such findings can then guide the development of effective strategies to boost resilience and mitigate the impacts of compounding risks on Chinese communities.

A fuzzy collaborative forecasting approach based on XAI applications for cycle time range estimation

Estimating the cycle time ranges of jobs is a critical task in a factory. this study proposes a fuzzy collaborative intelligence (FCI) approach to improve the precision of cycle time range estimation. In the proposed methodology, a DNN is first built to accurately predict the cycle time of a job. A random forest (RF) is then constructed to explain the DNN. Each decision tree of the RF is fuzzified to estimate the cycle time ranges of the jobs learned by the decision tree. A fuzzy collaboration mechanism is also established between decision trees to narrow the cycle time ranges. The proposed methodology is novel because the RF is not applied to predict job cycle times but used to explain and fuzzify the DNN without solving complex nonlinear programming problems. The proposed methodology has been applied to a real case. According to the experimental results, the FCI approach improved the estimation precision by 18%.

Hezbollah and Russian intelligence collaboration: a complex nexus

This paper delves into the intricate relationship between Hezbollah, the Shiite militant organization based in Lebanon, and Russian Intelligence Agencies (RIA). The collaboration between Hezbollah and RIA represents a complex interplay of geopolitical interests, regional dynamics, and strategic objectives. The paper examines the historical context of their relationship, the motivations behind their cooperation, the nature of their interactions, and the implications for regional and global security. Drawing upon a multidisciplinary approach that combines insights from Political Science, International Relations, Intelligence studies, and Middle Eastern studies, provides a comprehensive analysis of this collaboration. It also explores the broader implications for the Middle East, Russia's role in the region, and the challenges it poses to international efforts to counter terrorism and promote stability.

Privacy-Preserving Decision-Making Over Blockchain

Many blockchain applications require democratic on-chain decision-making. In this work, we propose a community-inclusive decentralised collaborative decision-making system with privacy assurance. Its key component is a two-stage voting scheme inspired by choice architecture. Our decision-making system is compatible with most existing blockchain infrastructures. In addition, it supports liquid democracy/delegative voting for better collaborative intelligence. Namely, stake holders can either vote directly on proposals or delegate their voting power to experts. When majority of voting committee members are honest, no one can derive voters' voting preferences or delegations with non-negligible probability. To support concurrent multiple voting events, we design a distributed batch key generation protocol that can generate multiple keys simultaneously by voting committee members with amortised communication cost of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathcal {O}(n)$</tex-math></inline-formula> per key, where <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$n$</tex-math></inline-formula> is the number of participants. Besides, our system supports “evolving committee”, i.e., voting committee members can be changed during the voting period. We implemented a pilot system in Scala, benchmark results indicate that our system can support large number of participants with high efficiency.

A heterogeneous fuzzy collaborative intelligence approach: Air quality monitor selection study

Smart backpacks with novel functions have an enormous potential market. Therefore, the ability to embed smart functions (e.g., air quality detection) into a backpack design is critical. This study aimed to select an air quality monitor suitable for the design of a smart backpack. Accordingly, a heterogeneous fuzzy collaborative intelligence approach is proposed. In the proposed methodology, experts apply heterogeneous methods to derive the fuzzy priorities of the criteria and evaluate the suitability of an air quality monitor. The evaluation results by all experts are aggregated using fuzzy-weighted intersection (FWI). After defuzzification, the top-performing air quality monitor is chosen. Based on the experimental results, the most critical features of an air quality monitor are price and size. Furthermore, not all air quality monitors are suitable for a smart backpack design. Among the compared air quality monitors, OO-KFR-PMA was the optimal choice. The contribution of this research is twofold. First, experts are enabled to apply heterogeneous fuzzy multi-criteria decision-making methods. Second, the unequal authority levels of experts are considered by aggregating their evaluation results using FWI.

Multi-robot geometric task-and-motion planning for collaborative manipulation tasks

We address multi-robot geometric task-and-motion planning (MR-GTAMP) problems in synchronous, monotone setups. The goal of the MR-GTAMP problem is to move objects with multiple robots to goal regions in the presence of other movable objects. We focus on collaborative manipulation tasks where the robots have to adopt intelligent collaboration strategies to be successful and effective, i.e., decide which robot should move which objects to which positions, and perform collaborative actions, such as handovers. To endow robots with these collaboration capabilities, we propose to first collect occlusion and reachability information for each robot by calling motion-planning algorithms. We then propose a method that uses the collected information to build a graph structure which captures the precedence of the manipulations of different objects and supports the implementation of a mixed-integer program to guide the search for highly effective collaborative task-and-motion plans. The search process for collaborative task-and-motion plans is based on a Monte-Carlo Tree Search (MCTS) exploration strategy to achieve exploration-exploitation balance. We evaluate our framework in two challenging MR-GTAMP domains and show that it outperforms two state-of-the-art baselines with respect to the planning time, the resulting plan length and the number of objects moved. We also show that our framework can be applied to underground mining operations where a robotic arm needs to coordinate with an autonomous roof bolter. We demonstrate plan execution in two roof-bolting scenarios both in simulation and on robots.

Frontiers of collaborative intelligence systems

The development of information technology has propelled technological reform in artificial intelligence (AI). To address the needs of diversified and complex applications, AI has been increasingly trending towards intelligent, collaborative, and systematized development across different levels and tasks. Research on intelligent, collaborative and systematized AI can be divided into three levels: micro, meso, and macro. Firstly, the micro-level collaboration is illustrated through the introduction of swarm intelligence collaborative methods related to individuals collaboration and decision variables collaboration. Secondly, the meso-level collaboration is discussed in terms of multi-task collaboration and multi-party collaboration. Thirdly, the macro-level collaboration is primarily in the context of intelligent collaborative systems, such as terrestrial-satellite collaboration, space-air-ground collaboration, space-air-ground-air collaboration, vehicle-road-cloud collaboration and end-edge-cloud collaboration. Finally, this paper provides prospects on the future development of relevant fields from the perspectives of the micro, meso, and macro levels.

Human–Artificial Intelligence Collaboration in Prediction: A Field Experiment in the Retail Industry

ABSTRACT This study investigates the role of human intervention in artificial intelligence/machine learning (AIML)-driven predictions. By doing so, we distinguish between three different types of human-AIML collaboration: automation, adjustable automation, and augmentation. We theorize that prediction uncertainty and time horizon represent two critical determinants of forecast accuracy. Based on a field experiment involving AIML-driven demand forecasts approximately 1,888 stock-keeping units in the retail industry, we rely on a multivalued treatment effect methodology to measure the effects of human-AIML collaboration on forecast accuracy. Our findings show that human intervention complements AIML-driven forecasts most effectively (augmentation) in predictions with long time horizons and low uncertainty. However human intervention is least likely to contribute to the effectiveness of AIML predictions (automation) in environments with short time horizons and high uncertainty. We discuss implications for extant theory and propose a framework outlining the conditions in which human intervention is most likely to add predictive value to human-AIML collaborations.

Cloud-assisted collaborative inference of convolutional neural networks for vision tasks on resource-constrained devices

This work analyses the most relevant research conducted under the mobile cloud computing paradigm to bring vision tasks supported by state-of-the-art deep convolutional neural networks closer to the end-user through collaborative intelligence. In particular, this review aims to comprehensively address collaborative inference on convolutional networks, offering the reader a detailed explanation of the main methods and technologies used to partition and deploy such models on the UE-edge-cloud continuum, which have made it possible to leverage the capabilities of resource-constrained devices to alleviate, and ideally eliminate, the traditional dependence on the cloud for high-performance computing, thereby enabling a more rational exploitation of the supporting hardware infrastructure. The paper details the technical aspects of the different frameworks designed to support these tasks, examining and comparing the mechanisms and techniques used to conceive pertinent configurations. Moreover, the study outlines the various algorithmic solutions developed for synthesizing optimal co-inference schemes, also covering the design conventions adopted and the tweaks implemented to optimize the overall performance delivered, concluding with a discussion of the specific challenges that have arisen thus far, as well as the following steps to be taken in this field.

Effects of Ransomware: Analysis, Challenges and Future Perspective

This review paper highlights the challenges and best practices in malware analysis, specifically focusing on the age of ransomware. It provides an overview of malware and its impact on computer systems and user privacy by lists various types of malware, including viruses, Trojans, spyware,adware, worms and highlights major malware attacks including the methods used and the resulting damages. Further, the article explores the challenges faced in ransomware analysis, including advanced encryption and evasion techniques, anti-analysis mechanisms, zero-day exploits and vulnerabilities, polymorphic and dynamic behavior, lack of resources, complexity of ransomware, collaboration difficulties, and cost implications. These challenges make it necessary for security researchers to constantly update their knowledge and techniques to effectively analyze ransomware. This study concludes best practices for ransomware analysis including isolating and segmenting ransomware samples in controlled environments, emphasizing behavior analysis and threat hunting, investing in advanced reverse engineering and automated analysis techniques, promoting collaborative intelligence and information sharing, and implementing security measures to protect against ransomware attacks. Additionally, the article briefly mentions static analysis techniques which explains that static analysis involves examining malware files and code without executing them. It can be used to identify ransomware characteristics, such as encryption algorithms, ransom demands, remote command execution, and obfuscation techniques. Moreover, file and code analysis methods, signature-based detection, code deobfuscation and unpacking techniques, and malicious document analysis and exploit detection are also suggested as part of static analysis.

Improving Biological Collections Data through Human-AI Collaboration

Biological collections play a crucial role in our understanding of biodiversity and inform research in areas such as biosecurity, conservation, human health and climate change. In recent years, the digitisation of biological specimen collections has emerged as a vital mechanism for preserving and facilitating access to these invaluable scientific datasets. However, the growing volume of specimens and associated data presents significant challenges for curation and data management. By leveraging human-Artificial Intelligence (AI) collaborations, we aim to transform the way biological collections are curated and managed, unlocking their full potential in addressing global challenges. We present our initial contribution to this field through the development of a software prototype to improve metadata extraction from digital specimen images in biological collections. The prototype provides an easy-to-use platform for collaborating with web-based AI services, such as Google Vision and OpenAI Generative Pre-trained Transformer (GPT) Large Language Models (LLM). We demonstrate its effectiveness when applied to herbarium and insect specimen images. Machine-human collaboration may occur at various points within the workflows and can significantly affect outcomes. Initial trials suggest that the visual display of AI model uncertainty could be useful during expert data curation. While much work remains to be done, our results indicate that collaboration between humans and AI models can significantly improve the digitisation rate of biological specimens and thereby enable faster global access to this vital data. Finally, we introduce our broader vision for improving biological collection curation and management using human-AI collaborative methods. We explore the rationale behind this approach and the potential benefits of adding AI-based assistants to collection teams. We also examine future possibilities and the concept of creating 'digital colleagues' for seamless collaboration between human and digital curators. This ‘collaborative intelligence’ will enable us to make better use of both human and machine capabilities to achieve the goal of unlocking and improving our use of these vital biodiversity data to tackle real-world problems.

AIS Effect on Human Behavior and Well-Being: An Empirical Analysis

The integration of artificial intelligence (AI) into various facets of our everyday existence has occurred at a rapid pace. This includes the presence of personal assistants on our smartphones as well as the algorithms that serve as the foundation for social media feeds. Artificial intelligence (AI) possesses the capacity to profoundly transform various aspects of human existence and professional endeavors. However, it is imperative to acknowledge that AI is not exempt from certain limitations and disadvantages. An aspect of considerable concern pertains to the potential psychological ramifications associated with the utilization of artificial intelligence (AI) technologies. With the escalating dependence on these technologies, there is a mounting apprehension regarding their potential adverse impact on human behavior and overall well-being. Further investigation is necessary in order to fully grasp the psychological ramifications of artificial intelligence, as these impacts are intricate and multidimensional in their entirety. The majority of tasks performed at the interface between humans and machines necessitate individuals to engage in novel and distinct activities, as well as to approach familiar tasks in alternative ways. The leaders of the future will be those who actively use collaborative intelligence, so revolutionizing their organizational practices, market strategies, industry dynamics, and, significantly, their workforce management. The field of artificial intelligence has witnessed significant advancements in various domains traditionally associated with human expertise, such as disease diagnosis, language translation, and customer service. Moreover, these advancements have been characterized by rapid progress. The emergence of such concerns is rooted in the realistic apprehension that artificial intelligence (AI) has the potential to gradually supplant human labor across all sectors of the economy. However, it is important to note that such a conclusion is not necessarily inevitable, or even the most probable. The current state of digital technologies exhibits an unprecedented level of responsiveness to human users, and conversely, humans have become increasingly attuned to utilizing these tools. The advent of AI will have a profound impact on the nature of work and the individuals involved in it. However, its primary effect will be to enhance and supplement human abilities, rather than displacing them entirely. The main aim to analyse the positive &amp; negative effects of AI on human behaviour &amp; well-being &amp; to understand how AI affects human behavior and well-being and deployment of AI technologies for the benefit of individuals and society.