Search Results (2,327)

Multi-agent path planning (MAPP) under partial observability requires agents to coordinate their movements and complete tasks efficiently without access to global information. The planning space and coordination complexity grow rapidly with increasing numbers of agents, targets, and obstacles. We formulate large-scale MAPP as a partially observable networked Markov decision process. Based on this formulation, we propose a decentralized model-based Actor-Critic using the Kronecker-factored trust region (DM-ACKTR) algorithm. The algorithm integrates local model learning with ACKTR-based policy optimization in an independent learning architecture. Each agent learns a local model to predict the next observation and reward. These predictions are used to construct additional transitions for Actor and Critic updates. A neighborhood-based communication mechanism incorporates information from nearby agents into value estimation. Region partitioning reduces each agent’s effective planning space. These improvements enable DM-ACKTR to continue outperforming the baseline algorithms as the scale of the MAPP problem increases. Experiments across three training and five evaluation scenarios show that DM-ACKTR achieves the best overall performance. Among the five evaluated algorithms, it consistently obtains the highest TCR and lowest CR, improving TCR by 2.06–4.35% and reducing CR by 11.26–25.95% relative to the respective best baselines. Full article

Tool-level attacks on Large Language Model (LLM) agents—poisoned tool descriptions, prompt injection, and capability misrepresentation—are universally effective, yet no existing defense provides comprehensive protection. We propose Architectural Intent Collapse (AIC), a formal framework capturing the systematic loss of communicative intent when text from heterogeneous sources is flattened into a single context window. Grounded as a novel instantiation of the Confused Deputy Problem, AIC reveals that the missing boundary is not permission but intent: the architecture cannot distinguish descriptive statements from prescriptive commands. We formalize AIC via an architectural collapse operator, introduce Intent Separation Degree (ISD) as a measurable metric, and develop a mechanism-based taxonomy of five intent-disguise attack types, including two previously undescribed (Conditional Latency and Inference Inducement). Experiments across 25 framework–model combinations (employing GPT-4o, Claude-4-Sonnet, Gemini-2.5-Pro, DeepSeek-V3, and Qwen3-32B as LLM backends) confirm that ISD degrades with description verbosity, strongly predicts defense effectiveness ($r=-0.97$), and is uniformly low across all current frameworks. Three root-cause defense principles are derived; one retains substantial protection against adaptive attackers. This research is useful for agent framework designers, security practitioners, and researchers seeking a principled understanding of why tool-level attacks succeed and how architectural defenses can address their root cause. Full article

Multipath routing in wireless sensor networks (WSNs) improves reliability by providing alternative forwarding paths when a route fails. However, mobile sinks make path maintenance difficult because sink movement can invalidate previously constructed source-to-sink routes. Existing protocols typically depend on either global path reconstruction, which increases control overhead, or footprint-chaining, which accumulates detours through previous sink positions and may weaken path independence. To address this problem, this paper proposes QL-LGMPRP, a reliability-aware local-grid-based multipath routing protocol that combines a sink-centered local grid, two-path delivery, link-quality-aware forwarding, and lightweight tabular Q-learning for waypoint adaptation. Mobility-related route changes are confined to the sink-centered grid, whereas a compact tabular Q-learning policy adjusts the primary-path direction using grid, link-quality, and energy-related state variables. The sink constructs a local grid around its current position, with cells sized to keep in-grid forwarding locally bounded. When an event occurs, the source computes an entry point on the grid perimeter and constructs two greedy paths: a primary path through a Q-learning-selected waypoint near the grid boundary and a backup path toward the current sink position. The Q-learning agent uses a compact tabular state representation that includes the boundary-cell index, residual-energy level, sink-grid position, and local link-quality information, and learns waypoint offsets using a reward that combines delivery success, transmission energy, and delay. This design confines routing adaptation to the sink-centered grid while allowing the waypoint policy to respond to heterogeneous link conditions. Simulation results under different sink speeds and interference conditions show that QL-LGMPRP maintains high delivery reliability while reducing detour-related forwarding costs relative to footprint-chaining and showing lower weak-link exposure than the geometric-forwarding comparison schemes. Full article

Large Language Models (LLMs) are increasingly integrated into agentic workflows that require extended reasoning, persistent state management, coordinated tool use, and controlled execution. As this operational scope expands, a central question emerges: whether probabilistic generation alone can reliably support coherent behavior across interacting system components. This paper addresses that question through a structural diagnostic review of contemporary agentic systems. Starting from LLM-based tutoring as an analytically demanding entry point and extending toward structurally related agent architectures, the paper draws on a five-phase review of $N=145$ research records. The analysis is organized through the Agentic Structure Taxonomy (AST), which structures the literature across four dimensions: Cognition, Interaction, Orchestration, and Governance. The review identifies five recurrent empirical problem patterns and uses them as abductive diagnostic cues for formulating seven cross-dimensional transition gaps that capture recurrent discontinuities at the boundaries between reasoning, state, control, and execution. From these gaps, fourteen structural constraints are derived across three control domains: state isolation, control alignment, and execution governance. These constraints are interpreted not as prescriptive design mandates, but as analytically derived conditions associated with reducing error propagation across subsystem transitions. The paper argues that reliability in agentic systems is shaped not only by model performance or prompt design, but also by whether the boundaries linking probabilistic reasoning to persistent state, orchestration, and execution are governed by explicit structural conditions. Full article

This study investigates the fixed-time quasi-consensus and quasi-containment control for multi-agent systems (MASs) under non-periodic unknown denial-of-service (DoS) attacks. Most available strategies fail to construct fixed-time observers and feasible corresponding parameter tuning rules to guarantee the precise fixed-time convergence of observer states to the convex hull trajectory spanned with multiple leaders under arbitrary non-periodic DoS interference. Moreover, most existing relevant fixed-time cooperative control methods for MASs commonly impose restrictive assumptions on system input matrices, requiring the matrix to be square and invertible, which severely limits their practical applicability. To overcome these limitations, the observers and corresponding parameter selection conditions are designed in this study, which can ensure that the observer states converge to the target trajectory formed by the leaders within a fixed time under non-periodic unknown DoS attacks. Then, based on the linear transformation of the state space and the theory of sliding mode control, a novel observer-based controller is proposed to solve the fixed-time quasi-consensus and quasi-containment control problems. The proposed approach remains effective even when, under mild conditions, the input matrix is non-square or non-invertible—a challenge that many existing methods cannot address. Finally, numerical simulations demonstrate that the proposed control strategy enables MASs with non-square input matrices suffering from unknown non-periodic DoS attacks to achieve fixed-time quasi-consensus and quasi-containment under mild conditions. Full article

Multi-agent reinforcement learning has emerged as a promising paradigm for cooperative unmanned aerial vehicle (UAV) swarm coordination. However, existing communication-aware MARL methods primarily focus on communication topology, message routing, and message aggregation, while the information content of the exchanged messages is often only implicitly controlled. In realistic UAV networks, inter-agent communication is constrained by limited bandwidth, communication range, energy consumption, and packet loss. It is therefore desirable for each UAV to transmit compact and task-relevant information rather than dense and redundant latent features. In this paper, we propose IB-CEMARL, an information-bottleneck-guided, communication-efficient multi-agent reinforcement learning framework for UAV swarms. We formulate inter-UAV communication as a minimal sufficient message-learning problem in which each UAV encodes its local observation into a stochastic bottleneck message before exchanging information with its neighbors. Cauchy–Schwarz divergence-based quadratic mutual information is adopted as a unified dependence measure to jointly regularize message compression, preserve decision-relevant information, and reduce statistical redundancy among neighboring UAV messages. Extensive experiments demonstrate that IB-CEMARL achieves superior cooperative performance, reduced message redundancy, and stronger robustness compared with representative communication-aware MARL baselines. In particular, IB-CEMARL improves the average return by $4.9\%$ and reduces inter-message dependence by $29.0\%$ compared with the KL-IB-MARL baseline while maintaining efficient communication under constrained bandwidth settings. Full article

Autonomous AI agents now hold funds, delegate authority to other agents, and transact at machine speed, while the governance apparatus meant to constrain them (policies, audits, compliance) remains documentation-based and limited by human latency. Better monitoring or filtering cannot close this mismatch: compliance must become a runtime, compositional, proof-carrying property of computation itself. We call the resulting discipline computational jurisprudence. This article is an integrative review of the four literatures the discipline must synthesize, namely, object-capability security; verifiable, proof-carrying, and zero-knowledge computation; policy-as-code and computational law; and agentic AI with its emerging payment protocols. Each supplies a mature mechanism the others lack, and none supplies a complete normative substrate. The synthesis is organized into three pillars: (i) a delegation calculus, under which authority can only attenuate as it propagates between agents, for which we prove monotone attenuation in the conjunctive caveat fragment and exhibit a counterexample outside it; (ii) runtime compliance proofs, a three-tier evidence regime (attested, optimistic, and zero-knowledge); and (iii) sealed delegation chains with graduated attribution, reconciling capability-based privacy with the accountability adjudication requires. A case study on agentic payments grounds the architecture and evaluates three components on two platforms, with five independent executions each: local capability verification against a centralized policy decision point, enforcement on the x402 payment path, and accumulator-based revocation. What the article offers is therefore a survey, a conceptual architecture with a formal core, and a partial evaluation of three components, not a fully implemented system; a status table marks that boundary component by component. Eight open problems define the research agenda. Full article

Unwanted telemarketing calls are a persistent source of consumer frustration and a legally regulated issue in Germany. At first glance, the idea of addressing this problem with AI-based voice technology appears contradictory: why should an automated caller help restore trust in a communication channel that has been damaged by aggressive outbound practices? This design-oriented case and prototype study argues that the paradox can be resolved through a different design logic. Rather than using AI to intensify persuasion, we present a role-based voice-agent architecture that constrains conversational behavior through narrow task boundaries, explicit escalation rules, and auditable data handling. The paper reports a transfer project involving FH Aachen students, Capgemini, and Fairdient GmbH. Methodologically, the work is positioned as a design-oriented case study with a prototype artifact. The contribution is threefold: first, we describe a three-agent architecture for outbound screening, consent-aware explanation, and inbound service; second, we derive governance principles for legally and ethically sensitive telephony, including transparency, bounded knowledge, privacy-preserving deployment, and human fallback; and third, we propose an evaluation framework covering conversion, compliance, hallucination control, user trust, and cost per validated outcome. The prototype does not yet claim large-scale field effectiveness. Instead, it offers a structured and empirically testable design for trustworthy voice automation in a domain where misuse, opacity, and user distrust are especially pronounced. Full article

Day-ahead dispatch of integrated energy systems (IESs) is commonly solved by mixed-integer linear programming when objectives, constraints, and forecasts are fully specified. In practice, however, operators often express temporary preferences in natural language, and day-ahead forecasts inevitably deviate from realized operation. These two conditions make a fixed optimization interface difficult to use without additional modeling effort. This paper proposes DARC (Decompose-Act-Repair-Critique), a constraint-diagnostic LLM agent framework for day-ahead dispatch of campus-level microgrid-type IES with under-specified operating requirements. DARC combines three language modules, namely a Decomposer for temporal structure, a Resolver for numerical schedule generation, and a Critic for root-cause diagnosis, with deterministic repair, constraint checking, and metric evaluation. In all main experiments, the Decomposer is instantiated by its deterministic rule-based variant for reproducibility, so the reported results reflect a loop with two LLM modules (Resolver and Critic). The Projector repairs candidate schedules where possible, while the Checker supplies mathematical facts that ground the Critic’s feedback. DARC does not model forecast uncertainty through sets or scenarios; robustness to forecast deviation is pursued operationally, through margin-aware repair and checker-grounded revision, and is assessed empirically. Experiments on a campus-level microgrid IES testbed show that DARC can incorporate natural-language preferences, improve realized feasibility under forecast noise relative to both a point-forecast MILP and an interval-robust MILP baseline in terms of strictly feasible configurations, at higher operating cost, and produce diagnostic feedback that is robustly more useful than an ungrounded LLM diagnoser under judge models from three families, and at least as useful as checker facts alone, with its advantage concentrated in actionable revision guidance. DARC is therefore positioned not as a replacement for MILP, robust formulations, or model predictive control on fully specified problems, but as a complementary interface for operational settings where preferences and forecasts are not completely formalized. Full article

Fast synthetic aperture radar (SAR) imaging simulation is required by many computer vision applications. Although the Shooting and Bouncing Ray (SBR) method has significantly accelerated electric field calculation, the number of ray tubes is still the bottleneck for SAR image simulation speed. This paper proposes an innovative adaptive SBR method driven by Q-learning for accelerated SAR imaging simulation. The core strategy is to convert the ray tube allocation into a reinforcement learning problem. The ray-shooting plane is dynamically partitioned into localized patches, where a Q-learning agent intelligently scales the ray density in real time. By observing the geometric features of the target surface, the agent learns to employ coarser ray tubes in flat regions to eliminate redundant computation, while deploying denser ray tubes in complex areas. A multi-objective reward function is designed to balance accuracy against computational resource consumption. Numerical experiments demonstrate that the proposed Q-learning-based SBR method drastically reduces computational cost while preserving imaging similarity. Full article

There is a rapidly rising demand for cosmeceutical solutions for skin lightening/brightening due to the growing need for dermatologic applications and consumers’ preference to enhance the brightness of skin. Traditionally, the main ingredients for developing dermatological products were the potentially harmful and controversial hydroquinone compounds; however, in recent years, the research field experienced a transition towards safer antioxidant-based agents. The present review analyzes the application and use of antioxidants in modern cosmetics as an innovative approach, and the chemical and biological mechanisms of their functions that could revolutionize the industrial applications. The most effective ones include competitive inhibition of tyrosinase enzyme, pheomelanin switch through the help of glutathione, protection from reactive oxygen species (ROS) in order to inhibit ultraviolet (UV)-stimulated production of melanin, as well as prevention of the migration of melanosomes using niacinamide. The review identifies and analyzes several components used as ingredients, such as vitamins C, Niacinamide, vitamin E, licorice root, epigallocatechin-3-gallate (EGCG), or resveratrol, among many more. In spite of their high efficiency, there are numerous problems associated with the instability, dermal irritation, and absorption of pure antioxidants. Finally, the legal framework developed by the Food and Drug Administration (FDA) and European Union (EU), potential risks like pro-oxidation damage, and future perspectives such as personalized skincare via machine learning and artificial intelligence (AI) are discussed. Full article

This paper addresses safe coverage control for multi-agent systems on a spherical surface with obstacles. While centroidal Voronoi tessellation has been widely used for distributed coverage generation, its extension to spherical environments with safety constraints remains relatively underexplored. To address this problem, we propose a spherical coverage control framework that integrates spherical centroidal Voronoi tessellation with control barrier functions. The proposed method assigns coverage objectives through spherical Voronoi partitioning and corresponding centroid computation, while safety is enforced through barrier-based constraints that prevent inter-agent collisions and obstacle collisions during motion. In addition, a surface-retention mechanism is incorporated to mitigate agent drift away from the desired spherical surface. The method is examined in a Python-based simulation environment with multiple obstacle layouts. The results demonstrate that the proposed framework enables safe distributed coverage over a spherical surface and that the surface-retention mechanism improves conformity of the agent trajectories to the target spherical manifold. The proposed approach provides a practical basis for safe multi-agent deployment in spherical environments. Full article

Wireless Sensor Network (WSN) underpins numerous applications, including environmental surveillance, industrial control, and smart cities, where reliable connectivity is vital for continuous sensing and data transmission. On the other hand, a small number of failed or compromised Critical Nodes (CNs) can badly fragment the network topology, interfere with communication, and impair performance. Consequently, devising effective solutions to the CN Detection Problem (CNDP) while taking connectivity, energy, and reliability into account remains a serious research endeavor. To tackle this challenge, this work proposes a Genetic Algorithm-assisted Damped Yo-Yo Grey Wolf Optimization framework with a Revitalized Boltzmann Distribution connectivity model (GA-DY-RBD-GWO). The CNDP is formulated as a node-elimination optimization problem that identifies the top-(k) CNs, where each search agent represents a candidate subset of k nodes. To realistically characterize network connectivity, a Revitalized Boltzmann Distribution (RBD)-based pairwise connectivity model is developed by jointly considering hop distance, residual path energy, and distance-based link reliability. Based on the resulting connectivity matrix, Total Pairwise Connectivity (TPC) is computed, and node criticality is quantified by the reduction in TPC after removing a candidate node set. To effectively explore the combinatorial search space, the Grey Wolf Optimization is augmented with a damped Yo-Yo control mechanism that adaptively balances exploration and exploitation during the optimization process. Furthermore, Genetic Algorithm-inspired crossover and mutation operators improve population diversity and avoid premature convergence, while elitist retention keeps the best-so-far candidate solution. By integrating realistic RBD-based connectivity modeling with an adaptive hybrid metaheuristic, GA-DY-RBD-GWO accurately identifies CNs whose deletion induces maximal TPC degradation. Extensive experiments under diverse network topologies, deployment scenarios, and spatial distributions demonstrate that the GA-DY-RBD-GWO exhibits superior performance over representative baselines, revealing that it is an efficient topology-aware solution for CNDP to improve the reliability of WSNs. Full article

Human–robot collaborative flexible job shop scheduling (HRC-FJSP) must coordinate heterogeneous capabilities, mode-dependent processing times, safety feasibility, and carbon constraints. The problem becomes harder when a collaboration mode that is attractive during planning becomes infeasible after a human enters the robot safety separation zone. Unlike conventional dynamic disturbances such as machine breakdown or order insertion, this event changes the feasible collaboration mode of the unfinished operation remainder rather than only delaying a resource or adding a job. This study formulates a carbon-aware dynamic HRC-FJSP and evaluates a carbon-aware multi-agent deep reinforcement learning scheduler (CA-MADRL) with local recovery after safety-proximity-induced collaboration disruption. The objective combines normalized makespan, carbon emission, and human workload imbalance with carbon accounting based on operation energy and time-varying grid carbon intensity. Across the benchmark cases, CA-MADRL obtains the best average global criterion (0.7235), wins nine of 12 cases, and achieves the lowest average carbon emissions among the compared policies (48.991 kg CO₂e). Sensitivity analysis shows that stronger carbon preference reduces emissions but increases makespan and tardiness, while adaptive collaboration outperforms fixed human–robot, human-only, and robot-only regimes. The results indicate that dynamic mode adaptation and local rescheduling improve carbon-aware collaborative schedules under safety disruption. Full article

In studies of intelligent agents, the pursuit–evasion problem, commonly related to the predator–prey paradigm, has been used as a reference setting for examining decision-making, coordination, and adaptation in multi-agent systems. In this paper, pursuit and evasion strategies are reviewed from their theoretical foundations to the gradual incorporation of adaptive methods based on evolutionary algorithms and machine learning. Analytical formulations drawn from control theory, game theory, and graph-based models are considered together with learning-oriented methods, including multi-agent reinforcement learning, neuroevolution, evolutionary robotics, and competitive coevolution. The literature is arranged by methodological paradigm so that the scope, limitations, and applicability of each approach can be discussed in relation to dynamic and uncertain environments. Through this organization, classical models, algorithmic developments, and recent research trends are brought into the same discussion, while relevant gaps and possible future directions in pursuit–evasion research are identified. The contribution of this work is a structured synthesis in which analytical and adaptive perspectives are brought together within a unified reference for researchers and practitioners in robotics, artificial intelligence, and multi-agent systems. Full article