Artificial Intelligence and Robotics Commons^™

Heterogeneous Multi-Robot Person-Following In Constrained Environments, Ori A. Miller

Electronic Theses and Dissertations

Maintaining visibility of a person requires effective systems. Security cameras or ground robots might be ideal, but they often fail in uncontrolled or unknown environments. A single ground robot struggles to navigate and track an agent at the same time. This work addresses the challenge by developing a multi-robot system with a slow ground robot and an agile aerial robot. Three methods are evaluated: FORWARD-PF, RL-Person Following (RL), and a baseline closed-loop method. FORWARD-PF proved the most reliable, completing all nine paths and reaching targets nearly twice as fast as RL. Despite completing seven paths, RL faltered on complex tasks. …

Supervised Classification Modeling On Louisiana Medicaid Data: A Comparative Study, Yead Rahman

Master's Theses

This thesis systematically optimizes and compares state-of-the-art supervised classification models for Louisiana Medicaid data targeting clinical services, COVID-19 infection, and tobacco use. These target variables are critically important as they represent key health outcomes and behaviors among Medicaid enrollees in Louisiana, a population often characterized by poverty and limited access to education. This study applies advanced machine learning techniques to identify the best model for multinomial and binary classification tasks. These include models such as Logistic Regression, XGBoost, AdaBoost, Random Forest, Decision Tree, Artificial Neural Networks, and Naïve Bayes. Extensive tuning of the hyperparameters and optimization of each classifier were …

Robotic Odor Source Localization Using Vision And Olfaction Sensing, Sunzid Hassan

Master's Theses

Robotic Odor Source Localization (ROSL) technology allows autonomous agents like robots to find an odor source in unknown environments. A successful odor source location depends crucially on an effective navigation algorithm that directs the robot towards the odor source. This thesis is a combination of three projects. First, we detail development of a versatile multi-modal robotic platform for ROSL real-world ROSL experimentation and discussed real-world validation of a traditional olfactionbased ROSL algorithm. Secondly, we introduced vision in ROSL by proposing a fusion navigation algorithm that integrates deep-learning enabled vision and olfaction-based navigation. This hybrid approach tackles challenges such as turbulent …

Democratization Of Custom, High Quality Large Language Models, Pablo Lopez

College of Computing and Digital Media Dissertations

Large Language Models (LLMs) have shown exceptional performance in several natural language processing (NLP) tasks. Customizing LLMs boosts their performance in domain specific tasks but typically requires substantial resources and effort for training, such as supervised fine-tuning. This research proposes methods to achieve significant accuracy improvements given minimal resources, particularly focusing on open-ended question answering with a given piece of context. We utilize an LLM’s self-generated training data to fine-tune the LLM and partial fine-tuning with on-demand GPU to reduce practitioner training costs. The research shows that these methods give significant performance gains in a Retrieval Augmented Generation (RAG) based …

The Impact Of Model Variations On The Robustness Of Deep Learning Models In Adversarial Settings, Firuz Juraev, Mohammed Abuhamad, Simon S. Woo, George K. Thiruvathukal, Tamer Abuhmed

Computer Science: Faculty Publications and Other Works

Rapid advancements of deep learning are accelerating adoption in a wide variety of applications, including safety-critical applications such as self-driving vehicles, drones, robots, and surveillance systems. These advancements include applying variations of sophisticated techniques that improve the performance of models. However, such models are not immune to adversarial manipulations, which can cause the system to misbehave and remain unnoticed by experts. The frequency of modifications to existing deep learning models necessitates thorough analysis to determine the impact on models’ robustness. In this work, we present an experimental evaluation of the effects of model modifications on deep learning model robustness using …

Developing Green Design, Leading Green And Low-Carbon Society, Yongxiang Lu

Bulletin of Chinese Academy of Sciences (Chinese Version)

Green design refers to a design principle and method that comprehensively considers energy and resource conservation, emission reduction, and environmental impact during the product manufacturing and operation process in the product design stage. It strives to reduce greenhouse gas emissions throughout the entire product lifecycle. In the era of information networks, green design is supported by big data, AI network collaborative design, network detection and monitoring, etc. It involves selecting energy-saving processes, green and low-carbon materials, and optimizing product geometry design and surface treatment to achieve efficient resource utilization and minimize waste. For example, selecting environmentally friendly materials through networked …

Challenges And Recommendations For Building Open Source Innovation Ecosystem For Large-Models In China, Xin Wen, Chao Zhang, Rui Guo, Kaihua Chen, Ze Feng, Qigang Zhu

Bulletin of Chinese Academy of Sciences (Chinese Version)

Addressing the current technological development issues that constrain the development of China’s large-scale model industry is needed to promote the continuous prosperity and development of the industry and enhance its international competitiveness. The study analyzes the significance of the open-source innovation ecosystem for the development of large-scale models in China. Based on reviewing the international experience of constructing the open-source innovation ecosystem, it further dissects the problems and challenges faced by the construction of the open-source innovation ecosystem for large-scale models in China and puts forward targeted suggestions. The study finds that the open-source innovation ecosystem for large-scale models in …

Research And Practice Of Digital Institute And Its System Framework, Jianjun Yu, Yue Wang, Kongmin Wang, Zhuomin Shi

Bulletin of Chinese Academy of Sciences (Chinese Version)

In the era of the digital economy, digital scientific research and digital government are both crucial components, serving as application scenarios enabled by new information technologies. Currently, the new generation of information technologies, especially artificial intelligence and big data, is instrumental in modernizing governance at scientific research institutions within Chinese Academy of Sciences (CAS). They notably drive paradigm shifts in scientific activities and accelerate the digital transformation of these institutions. The digital system formed by the digital transformation of management processes in scientific activities at research institutes is defined as the digital institute. This study analyzes the impact of digital …

Research And Implications Of The Us Clean Energy Strategy, Lanchun Li, Qing Liu, Wei Chen, Yun Tang, Jun Chen

Bulletin of Chinese Academy of Sciences (Chinese Version)

As the world enters a new period of carbon neutrality, the US government is actively building a clean energy innovation ecosystem through both internal and external measures. Systematically tracking and in-depth analysis of the intent, structure, approach, and other characteristics of the new phase of the US clean energy strategy is of practical significance to Chinese energy revolution. The US focuses on the strategic objectives of science and technology innovation, energy security, and infrastructure, and has constructed an innovation ecology characterized by technology lists, planning blueprints, full-chain research, and innovative subjects from the perspective of whole-government coordination, cross-institutional decision-making, deep …

Technology Governance And Governance Technology: From Perspective Of Regulatory Research On Blockchain Digital Assets, Yikai Wu, Guoan Li

Bulletin of Chinese Academy of Sciences (Chinese Version)

The Outline of the 14th Five-Year Plan takes blockchain as one of the key industries of the digital economy, and a number of ministries and commissions have also made clear deployments to accelerate the innovative application of blockchain, promote the digital transformation of the industry, and promote the high-quality development of the economy and society in the policy documents related to the informatization of the industry. The regulation and governance of blockchain digital assets cannot be separated from the understanding and analysis of blockchain technology itself, and observing the development mechanism of blockchain digital assets from the scientific and technological …

Ai-Driven Innovation And Development In Manufacturing: An Overview Of Trends, Issues, And Suggestions, Rongping Mu, Jingjing Guo, Yuchen Li, Qiang Li, Chun Jiang, Ze Feng, Guowei Dong

Bulletin of Chinese Academy of Sciences (Chinese Version)

Artificial intelligence technology has become a key force in driving innovation and development in global manufacturing industries, and its application in production automation, intelligent management, and other areas has become an important trend for innovation and development of manufacturing in all countries. This study comprehensively analyzes the innovation and development trend driven by artificial intelligence in global manufacturing, outlines the policies and measures of major countries in promoting the innovation and development of the manufacturing industries driven by artificial intelligence. It also points out the current situation and problems of the manufacturing’s innovation and development driven by artificial intelligence in …

Insights From Darpa’S Program Funding Layout In Artificial Intelligence, Zheng Su, Ning He, Qi Han, Qi Zhang, Xiaocheng Jiang

Bulletin of Chinese Academy of Sciences (Chinese Version)

Artificial intelligence, as a disruptor in the new wave of technological revolution and industrial transformation, had already become a focal point of strategic planning for the U.S. Defense Advanced Research Projects Agency (DARPA) as early as the 1960s. To explore the patterns of DARPA’s project funding allocation in the field of artificial intelligence, this study conducts an in-depth analysis of the agency’s budget reports from fiscal years from 2015 to 2025. The study identifies the following characteristics of DARPA’s project funding framework in the field of artificial intelligence: (1) distinctive application orientation; (2) highly flexible funding system; and (3) close …

Interventional Radiology's Exploration Into Artificial Intelligence, Raymond Nguyen

Master's Projects and Capstones

Background: Artificial intelligence (AI) has become more prominent in our daily lives in recent years. This includes various aspects of healthcare. Interventional radiology (IR) is one of these specialties that has taken strides in understanding how AI can be leveraged for patient care. This literature review aims to understand what areas will be most impacted by AI in IR and how it will influence both the patient and interventional radiologist.

Methods: Twenty-six publications from 2019-2024 were selected from PubMed and Scopus. Publications were sourced through a combination of keywords, subject headings (MeSH terms), and citation searching.

Results: This literature review …

Parallax-Tolerant Image Stitching With Geometric Structure Protection For Unmanned Ship Visual Perception, Zhilin Yang, Yong Yin, Rukai Zhang, Qianfeng Jing, Sen Jiang, Wenfeng Zhu

Journal of System Simulation

Abstract: In order to address the artifacts caused by misalignment in maritime image stitch with low texture and large parallax, a geometric structure parallax-tolerant image stitch algorithm based on point-line feature registration and optimal seam fusion is proposed. Line segment features are introduced into the traditional homography transformation based on point features, and potential coplanar local line segments are merged into global line segments to provide accurate alignment conditions for seam line fusion. In the image fusion stage, the energy function of seam cutting method is designed by using the color difference and gradient difference of tanh measure and introducing …

An Improved Cat Swarm Optimization For Heterogeneous Multiple Mobile Robots, Liang Kang, Yi Du, Lihua Yin

Journal of System Simulation

Abstract: At present, it is difficult to achieve the real homogeneity of the members of the multiple mobile robots. The existing swarm intelligence algorithm is also difficult to accommodate the heterogeneity of the team. Focusing on the heterogeneous cooperation of multiple mobile robots, the concept of mother and child robots is proposed. In order to realize the application of swarm intelligence algorithm in heterogeneous multiple mobile robots, the basic cat swarm algorithm is improved. The subdomain and neighborhood of cat swarm are defined, and eight improvements of cat swarm algorithm are proposed, including priority of search direction, extended trajectory tracking …

Research On Path Optimization Algorithm In Dynamic Routing Environment, Xin Xie, Xiaobing Hu, Hang Zhou

Journal of System Simulation

Abstract: In a real dynamic routing environment, static path optimization (SPO) and traditional dynamic path optimization (DPO) tend to encounter issues such as detours, reversals and high computational complexity due to frequent real-time optimization calculation. To address these problems, a novel restart co-evolutionary path optimization (RCEPO) method based on the ripple-spreading algorithm (RSA) is proposed. This method integrates the path optimization process with the dynamic changes of the routing network environment to enhance the effectiveness of path optimization. Moreover, the path reoptimization calculation is performed only when the dynamic changes in the routing environment exceed the predicted range, thereby reducing …

Parametric Identification Of Ship Maneuvering Motion Response Model Based On Square Root Cubature Kalman Filtering, Qinghao Li, Junsheng Ren, Yan Hua

Journal of System Simulation

Abstract: The system identification algorithm, based on the square root cubature Kalman filter (SRCKF) is proposed to address issues such as low accuracy, poor robustness, and weak generalization ability encountered by the extended Kalman filter (EKF) algorithm in parameters identification of ship maneuvering motion models. This algorithm, within the framework of CKF, replaces the original covariance matrix with its root mean square and utilizes triangular decomposition for prediction and update to enhance identification stability. The EKF is used as a comparison algorithm to identify the parameters of the second-order nonlinear response model of a ship with rudder angles that comply …

A Method Based On Deep Learning For Assisting Sins/Dvl Integrated Navigation, Xinghong Kuang, Aowei Huang

Journal of System Simulation

Abstract: The navigation and positioning accuracy of an Autonomous Underwater Vehicle (AUV) affects the efficiency of the AUV to a certain extent, and since GNSS cannot be used underwater, the integrated navigation system of Strapdown Inertial Navigation System/ Doppler Velocity Log (SINS/DVL) has been widely favored. However, DVL will fail in some cases, and if DVL is isolated directly, the system will become a pure inertial navigation system, which seriously affects the accuracy of navigation and positioning. In order to cope with the situation that DVL is missing in some beams, a DLinear-Informer assisted integrated navigation algorithm is proposed. Through …

Dynamic Data Driven Simulation: An Overview, Xu Xie, Xiaogang Qiu, Yizheng Bao, Kai Xu

Journal of System Simulation

Abstract: Dynamic data driven simulation is a simulation paradigm which integrates simulation and data together. This paradigm continuously feeds real-time data into the simulation, enabling the simulation be dynamically adjusted by the data, which thus improves the simulation-based estimation and prediction capability. Due to this integration, the dynamic data driven simulation can estimate system states and predict future state evolution more accurately. This paper reviews the origins and basic concept of dynamic data driven simulation, and introduces several simulation paradigms originated from the idea of "integrating models with data", and identifies the linkages and differences among them. The particle filterbased …

Study On Robust Chance Constrained Optimization Of Multi-Energy Supply System Based On Wind And Solar Power Combined Output Simulation, Zhe Bao, Wei Li, Xiaofang Zhang, Zongyuan An, Ye Xu

Journal of System Simulation

Abstract: In order to effectively avoid potential imbalance between supply and demand caused by the uncertainty of wind and solar power outputs, and promote the sustained development of the multi-energy supply system, a robust chance-constrained optimization model is developed for identifying optimal operation strategies under complexities and uncertainties through incorporating Copula theory, chance-constrained programming, and robust programming within a general framework. The results show that this model can not only accurately characterize the distribution probability of combined outputs of wind and solar power and formulate the operational strategies under low default risk conditions, but also reduce the proportion of highrisk …