The rapid rise of large-scale foundation models and autonomous AI agents is reshaping the knowledge ecosystem. Academic journals, long designed for human-to-human communication, now face a growing mismatch: an increasing share of knowledge is consumed by machines, not only by people. This editorial argues that this shift calls for a move toward a third generation of academic journals enabled by parallel publishing. Using the historical transition from hunting and gathering to agriculture as an analogy, we suggest that knowledge production must evolve from passive harvesting to systematic cultivation of content that is reliable, structured, and usable by models and agents. We then examine the risks of low-quality, fabricated, and citation-manipulated publications, and discuss how the role of researchers is changing from authors to stewards of reusable knowledge assets. Finally, we draw attention to recent community efforts promoting parallel embodied and digital publishing systems and propose knowledge asset tokenization as a practical mechanism to improve traceability, incentivize quality, and support sustainable human-AI co-evolution.

Robotic autonomous artistic creation represents an emerging frontier at the intersection of artificial intelligence (AI) and human creativity, pushing embodied AI systems toward more human-like perceptual, cognitive, and decision-making capabilities. However, the artistic creation process inherently involves complex action sequences, intricate motion trajectories, and subtle aesthetic representations, presenting substantial challenges for robotic modeling and execution. This paper proposes a vision-language-action (VLA) driven robotic painting framework that introduces advanced VLA models into the Parallel Art architecture to provide a more effective and systematic solution to these challenges. Building upon parallel systems theory and the ACP (artificial systems, computational experiments, and parallel execution) methodology, the framework integrates VLA models into the process of artificial systems, computational experiments, and parallel execution. It constructs virtual artistic creation environments for efficient model training, develops VLA-based artistic creation models through iterative computational experiments, and enables autonomous painting perception and decision-making through creative virtual-real interaction, thereby forming an integrated workflow for robotic artistic creation. The robotic painting results demonstrate the effectiveness of the proposed method and highlight its advantages over the existing approaches.

The problem of trajectory tracking for quadrotor unmanned aerial vehicles (UAVs) is investigated in this paper. An iterative linear quadratic regulator (iLQR) based model predictive control (MPC) strategy is proposed. The proposed iLQR-MPC strategy solves the nonlinear optimal control problem using the iLQR algorithm and implements the control in a receding horizon manner. A constrained iLQR algorithm is designed in the framework of the augmented Lagrangian method to solve the optimization problem induced by the trajectory tracking. Finally, a comparative simulation experiment is conducted to demonstrate the effectiveness and advantage of the proposed control strategy.

In this paper, a fixed-time optimized control scheme with prescribed performance is presented for bipartite consensus of nonlinear multi-agent systems (MASs). To enforce predefined transient and steady-state performance within a fixed time, a fixed-time prescribed performance function (FTPPF) is introduced. The control design is achieved via an identifier-actor-critic framework that approximates the solution to the Hamilton-Jacobi-Bellman equation. Unknown system dynamics are compensated by fuzzy logic systems, and a rounding quantizer processes the signal of the leader to reduce the communication costs. The proposed method ensures all consensus errors converge with prescribed performance bounds in a fixed time. Simulation results validate the theoretical findings.

A model-free adaptive heading control strategy for unmanned surface vessels (USVs) is proposed by integrating extended state observer (ESO) with model-free adaptive control. To handle internal model perturbations and external disturbances, a discrete linear ESO is employed to estimate the aggregate uncertainties. Then, the model-free adaptive heading controller is constructed by utilizing system input and output data, and the estimation information deriving from the ESO. Furthermore, a dynamic event-triggered mechanism (ETM) is designed and embedded in the controller to save communication data. Compared with the periodic sampling, the introduction of the dynamic ETM enables the controller to achieve a comparable control performance with only 5.35% of the communication frequency. Simulation comparison and analysis exhibit that the proposed heading controller can achieve superior heading control effect for USVs and show broader application prospect.

This study explores the design of an unknown input observer for discrete-time nonlinear Markov jump systems based on Lyapunov theory and Takagi-Sugeno fuzzy models. Firstly, the nonlinear system is transformed into a linear system using linear system theory and the Takagi-Sugeno fuzzy model is applied to solve the nonlinear problem. Then, under the conditions where unknown inputs exist in the state equations of the nonlinear Markov jump system, an unknown input observer is proposed so that the error system for state estimation has no relation to the unknown inputs. The feasibility conditions of the unknown input observer steady-state estimation error system are derived to ensure that the estimation error can approach zero. Additionally, linear matrix inequalities are used to address the feasibility problem of unknown parameters. The introduction of relaxation matrices reduces the conservativeness of the conditions. Finally, the proposed unknown input observer is validated using the vehicle lateral dynamics system. The results indicate that the unknown input observer can accurately track the actual state of the system, and the corresponding estimation errors converge to zero.

We present a safety-critical finite-time control framework for sampled-data multi-robot system coordination. We formally define and construct a finite-time discrete control Lyapunov function (FT-DCLF) and derive sufficient conditions that ensure convergence within a preset number of sampling steps, thereby enhancing both applicability and convergence speed. In parallel, we improve upon a discrete control barrier function (DCBF) constraint. This constraint addresses the continuous-discrete mismatch (“safety gap”) and ensures all-time safety, while mitigating deadlock and resolving performance-safety conflicts that are common in conventional DCBFs for obstacle avoidance. Both components, along with input bounds, are integrated into a single quadratically constrained quadratic program (QCQP), augmented with feasibility-aiding slack variables for real-time implementation. Simulation results demonstrate that the proposed method outperforms conventional DCLF-DCBF approaches.

This paper investigates platoon control under denial-of-service (DoS) attacks. Unlike conventional formations, this study primarily focuses on leaderless vehicle platoons and develops a corresponding model. Since communication among vehicles in the platoon is not only disrupted by DoS attacks but also constrained by limited network resources, a resilient dynamic event-triggered mechanism is proposed. Based on this mechanism, an event-triggered distributed formation control strategy is developed. Furthermore, a Lyapunov function is constructed, which provides a rigorous proof of system stability and a scheme for controller gain design. Finally, a simulation of a four-vehicle platoon model under DoS attacks based on sampled-data event-triggered is designed to further verify the effectiveness of the proposed control strategy and the stability of the vehicle platoon system.

This paper addresses sequential synchronization for first-order affine systems in the presence of bounded disturbances. We develop a reinforcement learning (RL) based controller that combines an anisotropic weighted power-sign baseline with a projected residual actor-critic. The baseline uses a diagonal weighting to shape the direction field of a transformed error and thereby encode the desired convergence order across state channels. A lightweight projection comprising a norm cap and a half-space alignment with the descent direction filters the residual policy so as to preserve the baseline Lyapunov decrease. We establish fixed-time stability and sequential convergence for the baseline via a composite Lyapunov argument, and prove that the RL residual maintains these properties under the projection. A spacecraft translational tracking case study with bounded sinusoidal disturbances validates the design: the position states converge in the prescribed order, the residual actions remain consistently bounded with respect to the baseline control magnitude and direction, and learning signals stabilize. Quantitatively, axis-wise first-hit times, control energy, and peak control corroborate the theoretical guarantees while illustrating performance benefits of the residual layer.

A systematic analysis and investigation of the electromagnetic field characteristics in radial magnetic levitation bearings is conducted using the finite element simulation methodology. A mathematical model of electromagnetic force for a single magnetic pole is established to derive the relationship between electromagnetic force and parameters such as current and air gap. Finite element simulations conducted via COMSOL multiphysics yield results that are consistent with theoretical calculations, while further analyzing magnetic field distribution, air gap flux density, and electromagnetic force characteristics under saturation conditions, thereby effectively compensating for errors in theoretical computations. This study demonstrates the effectiveness of the finite element method in the design and optimization of magnetic levitation bearings, providing a more precise and reliable analytical approach for practical engineering applications.

Artificial intelligence (AI) agents driven by large language models have demonstrated substantial advancements in addressing complex problem-solving tasks. While current AI agents’ collaborations are predominantly organized by centralized platforms, this organizational model inherently introduces systemic challenges such as data monopolies and interoperability barriers, thereby hindering cross-domain collaborative potential. Although current multi-agent collaboration methods have enabled secure data and resource sharing, they still fall short in utilizing distributed knowledge and adopting validatable collaborative workflows. To overcome these limitations, this paper proposes a distributed multi-agent collaboration mechanism leveraging the smart-contracts-based validation method with specifically designed planning, execution and validation smart contracts for robust workflow validation. An illustrative example of decentralized science collaborative writing is provided to demonstrate the applicability of the proposed mechanism. To further bridge the data silos and autonomously update the smart contracts, we employ decentralized autonomous organizations and operations (DAOs) and adopt an auction-based tool supplier matching model to architect the organizing and incentive scheme to enhance self-adaptivity and scalability. To further evaluate the performance of the proposed mechanism, we delineate a hierarchical indicator system and discuss the corresponding key technologies. This paper outlines a smart-contract-based approach to enhance the reliability of multi-agent collaboration, which could serve as a methodological basis for constructing future distributed multi-agent collaboration ecosystems.