Adaptable System for Disaggregated Distributed AI Chat Enablement (D2ACE) to Support Mission Engineering

In this paper, we begin our process of incorporating an AI bot in an online chat tutoring setting as a support for the tutor. We explore how an AI bot could give suggestions for tutor messages, although the human tutor will control how to communicate with the student. Tutoring, an important dimension of networked learning, has long been seen as a beneficial approach to students’ learning. An AI bot has the potential to aid tutors in the tutoring process and contribute to the scalability. The present pilot study was conducted in the tutoring setting of the Math Coach program. In the program, teacher students aid students from upper primary school to upper secondary school in mathematics through an online text-based chat system. Llama2 was used as a large language model (LLM), fine-tuned for Swedish comprehension utilizing the Math Coach system's chat logs. Four coaches, teacher students at a technical Swedish university and active in the Math Coach program, were invited to interact with the AI bot and participate in a group discussion. The coaches interacted individually with the AI bot while the chat conversation was displayed on a monitor so all participants could discuss the interaction while it took place. A semi-structured interview approach was taken and the participants were also encouraged to 'think aloud' about their experience. In the discussions, the coaches expressed surprise by the AI's social aspect. They perceived the AI bot as friendly with a positive attitude and were especially surprised by its ability to correctly place appropriate emojis. The coaches agreed that the AI was able to ask both appropriate and helpful questions and share some good guidance for how to proceed in the problem-solving process. However, they felt that the AI bot was not able to offer sufficient mathematical guidance, oftentimes the AI bot was confidently wrong. It also wrote too long messages, which humans would typically separate into several chat messages, and did not wait for a response but instead moved too quickly towards the solution. Moving forward we plan to address the effects of improved prompts on the AI bot and continue finetuning the LLM. We will continue to conduct pilot studies and eventually conduct more large-scale empirical studies.

When we speak of the study of discourse and interaction, we are mostly concerned with the analysis of synchronous conversation as opposed to written texts, although we might also consider text‐based computer‐mediated communication. This is not to say that writing is not interactive: people write letters to each other and, more recently, engage in asynchronous written interaction through blogs and social networking sites. The focus of the analysis of discourse and interaction, however, has traditionally been on “real‐time ”interaction in which participants are temporally copresent (though they may be spatially distant) and jointly engaged in some kind of social action. Examples include face‐to‐face interactions of various kinds (from religious rituals to casual conversations), telephone conversations, video conferences, chat and instant messaging sessions, and the interaction between avatars in online games or in virtual worlds. Discourse analysts are chiefly concerned with how parties in interaction work together to construct their social identities, their social relationships and their social realities moment by moment through language.

Number of instant messaging (chat) applications grows rapidly, not only in everyday communication of ordinary people, but in business correspondence. Consequently, security and privacy of these kinds of applications are very important. In this paper, CryptoCloak protocol for secure and private chat communication is introduced, together with the prototype IM application implemented in Java programming language that illustrates how the CryptoCloak protocol can be used in practice.

Food industry and engineering—Quo vadis?

Collective autonomic systems (CAS) are adaptive, open-ended, highly parallel, interactive and distributed software systems. They consist of many collaborative entities that manage their own knowledge and processes. CAS present many engineering challenges, such as awareness of the environmental situation, performing suitable and adequate adaptations in response to environmental changes, or preserving adaptations over system updates and modifications. Recent research has proposed initial solutions to some of these challenges, but many of the difficult questions remain unanswered and will open up a rich field of future research.

Currently, there is a lack of tools to build autonomous software systems in a pervasive and distributed context. In this paper, we present a model and a development tool called Agentphi for building flexible agent architectures from fine-grained components. Reusable components implement non-functional mechanisms such as communication, mobility or adaptation skills. Each agent can dynamically and autonomously change its components to suit its runtime context, improving safety and performances of applications, particularly in open, pervasive or large-scale distributed applications. We also show how to use our agents to build applications from autonomous and adaptable systems.

Recent breakthroughs in robotics and artificial intelligence have permitted more autonomous systems, although most current techniques are limited by inflexible control structures developed from classical automation. Such frameworks stand in stark contrast to biological systems, which develop intelligence via ongoing structural and functional reconfiguration. Neural morphogenesis takes a developmental approach to machine intelligence, seeing robotic cognition as a dynamic and adaptable process. In this paradigm, artificial agents gradually adapt their internal circuitry, behavioral tactics, and physical morphology as they interact with their surroundings. Learning and design therefore occur together rather than sequentially. The approach integrates embodied perception and safety-governed developmental adaptation within a closed-loop control framework. We introduce a three-layer developmental controller that couples online neural plasticity, morphogenetic regulation, and energy-aware policy adaptation under explicit rollback safety constraints. In over 105 simulation cycles involving populations of 50, 100, and 200 agents, neural morphogenetic architectures achieved energy efficiency gains of up to 47%, network modularity increases of around 40%, and reductions in informational entropy between 13 and 15% (p < 0.01) when compared to deep reinforcement learning and model predictive control baselines. These gains were accompanied by improved cooperative behavior, steady performance under stress, and greater resistance to environmental shocks. Preliminary bio-hybrid tests suggest a link between morphological flexibility and integrated information (Φ), emphasizing the impact of physical structure on cognitive capabilities. These findings establish brain morphogenesis as a strong basis for adaptable, resilient, and sustainable intelligent robotic systems. Received: 5 September 2025 | Revised: 16 December 2025 | Accepted: 26 February 2026 Conflicts of Interest The author declares that he has no conflicts of interest to this work. Data Availability Statement The data that support the findings of this study are openly available in Zenodo at https://zenodo.org/records/18175173, reference number [45]. Author Contribution Statement Edwin Gerardo Acuña Acuña: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.

Exclamation mark or kantanfu in Japanese, is one of written language elements used to convey expressions. Based on exclamation mark usage, the meaning and nuance of a sentence could differ. Amidst the COVID19 pandemic that limits one’s interaction to another, the frequency of digital communication has significantly increased. Digital communication through chats has become an activity that cannot be separated from everyday life. Therefore, it is important to understand exclamation marks correctly. The aim of this study is to avoid misunderstandings that could happen through chats in Japanese. This study will compare the use and interpretation of exclamation marks in Japanese chat between Indonesians and Japanese. Japanese chat messages containing exclamation marks were analyzed using Hymes’ SPEAKING theory. This study uses Creswell's convergent mixed methods, with quantitative data of Indonesians’ questionnaire survey results, and qualitative data of interviews with Japanese people. Study results show that there are differences in the use and function of exclamation marks between Indonesians and Japanese.

This study investigated the impact of undergraduate learning assistants on students’ educational experiences in large lecture neuroscience courses. These courses were highly structured and incorporated weekly assignments co-developed with undergraduates trained in a Certified Learning Assistants Program (CLAP). We employed four undergraduate learning assistants (LAs) in two large lecture general education neurobiology courses who assisted the instructor in developing targeted homework assignments based on lecture content. Students were encouraged to interact with LAs via email, Canvas, or an online messaging platform. The messaging platform did not include the instructor or graduate student teaching assistant (TA). LAs provided weekly in-person homework assignment introductions and review sessions during lectures. At the end of the course, students answered surveys about their learning experience, communication preferences, and overall sense of belonging. Findings suggest that the incorporation of undergraduate learning assistants improved students’ sense of belonging and confidence in large lecture neuroscience courses. Students were more comfortable asking content questions of LAs in a collaborative group messaging chat without graduate TAs or the Instructor. Group chat communications were preferable to using their course Canvas site or traditional email. Student interest in the course was increased by knowing that the course assignments were generated by the undergraduate LAs. Although students felt that LAs improved their sense of belonging, they were still less comfortable interacting with LAs in person compared to online platforms. In summary, LA involvement provided effective peer support for students and student perspective for the instructor, increasing students’ sense of belonging and interest in course content.

Risk management is a well-known method to face technological challenges through a win–win combination of protective and proactive approaches, fostering the collaboration of operators, researchers, regulators, and industries for the exploitation of new markets. In the field of autonomous and unmanned aerial systems, or UAS, a considerable amount of work has been devoted to risk analysis, the generation of ground risk maps, and ground risk assessment by estimating the fatality rate. The paper aims to expand this approach with a tool for managing data protection risks raised by drones through the design of flight maps. The tool should allow UAS operators choosing the best air corridor for their drones based on the so-called privacy by design principle pursuant to Article 25 of the EU data protection regulation, the GDPR. Among the manifold applications of this approach, the design of fly zones for drones can be tailored for public authorities in the phase of authorization of new operations, much as for national Data Protection authorities that have to control the lawfulness of personal data processing by UAS operations. The overall aim is to present the first win–win approach to data protection issues, aerospace engineering challenges, and risk management methods for the threats posed by this technology.

Autonomous vehicle perception systems require robust pedestrian detection, particularly on geometrically complex roadways like Type-S curved surfaces, where standard RGB camera-based methods face limitations. This paper introduces YOLO-APD, a novel deep learning architecture enhancing the YOLOv8 framework specifically for this challenge. YOLO-APD integrates several key architectural modifications: a parameter-free SimAM attention mechanism, computationally efficient C3Ghost modules, a novel SimSPPF module for enhanced multi-scale feature pooling, the Mish activation function for improved optimization, and an Intelligent Gather & Distribute (IGD) module for superior feature fusion in the network's neck. The concept of leveraging vehicle steering dynamics for adaptive region-of-interest processing is also presented. Comprehensive evaluations on a custom CARLA dataset simulating complex scenarios demonstrate that YOLO-APD achieves state-of-the-art detection accuracy, reaching 77.7% mAP@0.5:0.95 and exceptional pedestrian recall exceeding 96%, significantly outperforming baseline models, including YOLOv8. Furthermore, it maintains real-time processing capabilities at 100 FPS, showcasing a superior balance between accuracy and efficiency. Ablation studies validate the synergistic contribution of each integrated component. Evaluation on the KITTI dataset confirms the architecture's potential while highlighting the need for domain adaptation. This research advances the development of highly accurate, efficient, and adaptable perception systems based on cost-effective sensors, contributing to enhanced safety and reliability for autonomous navigation in challenging, less-structured driving environments.

The integration of machine learning (ML) and artificial intelligence (AI) into manufacturing has transformed traditional processes, creating smarter, more efficient, and adaptable production systems in line with Industry 4.0 principles. Advanced AI applications can enhance automated quality control by detecting defects in real time, thus improving product reliability and reducing waste. Technologies like digital twins and generative design allow manufacturers to simulate, optimize, and innovate production processes before physical implementation, cutting down on development time and costs. In supply chain optimization, ML algorithms enhance inventory management, demand forecasting, and logistics efficiency. The benefits are substantial are challenges such as big data management, integration with legacy systems, workforce reskilling, and ethical concerns. Future trends include the emergence of autonomous manufacturing systems, sustainable AI applications, and improved interoperability through the Internet of Things (IoT) and edge computing in the manufacturing sector.

In the evolving realm of autonomous vehicle navigation, the integration of fuzzy logic and neural networks presents a formidable challenge, particularly in the context of real-time, on-the-fly neural network training. This paper addresses the gap in dynamic and adaptable training methods necessary for navigating unpredictable environments with limited computational resources. The primary objective of our study is to empirically validate a hybrid training approach that combines fuzzy logic with back-propagation learning algorithms, aiming to optimize neural network performance under hardware constraints. Our methodology leverages a fuzzy logic trainer to provide initial training sets dynamically, which guide the neural network in adjusting its weights in real time, thus facilitating adaptive learning during navigation tasks. The findings reveal that this integrated approach not only enhances the learning efficiency of neural networks but also significantly improves navigation accuracy in real-time scenarios. These advancements contribute to the field by demonstrating the feasibility of deploying more adaptable and robust autonomous navigation systems, potentially expanding their application in more diverse and challenging environments.

The article deals with the pragmalinguistic research of English internet relay chat as a type of mass computer mediated communication in which disembodied distant interlocutors exchange text messages online. The main aim of this study is to identify the structural, formal and functional features of the English chat, which determines the corresponding research objectives: to describe chat-communication as the implementation of a phatic function in the language; to determine the specifics of its anthropocentre. Thus, the study focuses on one of the components of internet chat communication, namely, its anthropocentre. Chat communication is seen as polylogical interaction with different chatters initiating interaction technically on end. Our research found that phatic motivation and absence of face-to-face contact of chatters alongside with fast tempo of communication and short period of text product existence create its specific structural and language features. Chat communication is seen as a prototypical type of phatic computer mediated communication with the help of which chat users realize the playful behaviour impulse of communication, ambition to experience emotionally rich and carefree chatting. The research sample comprised chat logs that saw decency violation both in terms of topics discussed and modes of expression. Emotionally charged obscene vocabulary was vastly used in a number of chat logs, too. The research into the reasons of it showed that chatters resorted to obscene talk in order to get rid of emotional stress and psychological frustration. Chat discourse is largely composed of phatic and metacommunicative remarks where all chatters can enjoy the function of both the addresser and the addresee. A number of logs saw practically no exchange of factual information.

The NASA Engineering and Safety Center (NESC) was established in 2003 (after the Columbia accident) to provide an independent technical resource for the resolution of challenging technical problems (through the use of studies, analysis, tests, etc.) for NASA programs and projects. Since its inception, NESC has completed nearly 1000 technical assessments for NASA’s Human Exploration and Operation Mission Directorate (HEOMD), Science Mission Directorate (SMD), Space Technology Mission Directorate (STMD), and Aeronautics Research Mission Directorate (ARMD). Of the SMD related assessments, several were for the resolution of technical problems, analysis, or studies related to NASA’s astrophysics missions in various phases of the project from design to operation. Some of the recent examples of NESC technical support for NASA astrophysics missions have been for: Hubble Space Telescope (HST), Chandra X-ray Observatory (CXO), Fermi Gamma-ray Space Telescope, Kepler Space Telescope, Transiting Exoplanet Survey Satellite (TESS), James Webb Space Telescope (JWST), and Laser Interferometer Space Antenna (LISA). In this paper, we outline some of the technical challenges faced by these astrophysics missions and describe how NESC contributed to their resolution. The case studies cover a wide range of disciplines involving space telescopes, detectors, lasers, and attitude control systems. These efforts include innovative solutions for extending the life of the missions, technical resolution of challenging problems, strategies for risk mitigation, and failure investigations combined with lessons learned reports to advance discipline knowledge, enhance NASA capabilities, and avoid future problems.