Do you speak my language? Knowledge challenges in open environments

Internet auctions in open dynamic environments have been attracting increasing attention. We analyze with a bottom-up approach the competition between artificial intelligent agents in Continuous Double Auction markets. In almost all previous works agents have a fixed bidding strategy during the auction, usually under static symmetric environments. In our simulations we allow the soft-agents to learn not only about how much they should bid or ask, but also about possible switching between the alternative strategies. We examine the behaviour of strategic traders under dynamic asymmetric environments thus extending previous results. This analysis is important in the design and performance of auctions in the real world (stock exchanges, commodity markets, emission permits, and B2B exchanges) and in the applications of auction theory to many problems in management and production, far beyond market design (market oriented programming).

Multi-agent swarm collaboration, inspired by the collective behaviors of biological swarms in nature, has wide applications in dynamic open environments. However, hybrid action spaces in multi-agent reinforcement learning (MARL) present a critical challenge: the inherent coupling between discrete and continuous actions severely undermines policy stability and convergence, especially under dynamic topologies. Existing methods fail to decouple this coupling, leading to suboptimal policies and unstable training. This paper addresses the core problem of action coupling under dynamic topologies, proposing a Parametrized Graph Convolution Reinforcement Learning (P-DGN) method. Operating within the actor-critic framework, P-DGN decouples the optimization pathways for hybrid actions, with a biomimetic observation design inspired by starling flock behaviors: each agent only observes the states of its seven nearest neighbors to achieve efficient local interaction and global collaboration. Its actor network uses multi-head attention to build dynamic relation kernels, develops temporal relation regularization (TRR) to improve policy consistency across time steps, and generates continuous actions with a Gaussian policy. Meanwhile, P-DGN's critic network, based on deep Q-network (DQN), evaluates Q-values for discrete actions to guide optimal choices. We evaluate P-DGN in two different multi-agent cooperative environments. Experimental results show that compared with parametrized deep Q-network (P-DQN) and DQN baseline, the proposed method has faster convergence speed and stronger training stability. Moreover, with dense rewards, P-DGN agents learn emergent tactics like encirclement. Overall, P-DGN offers a new approach for optimizing hybrid action spaces in multi-agent systems within open, dynamic environments, balancing theoretical generality with practical utility, and its biomimetic design provides a biologically plausible framework for multi-agent swarm collaboration.

Adaptive team performance is essential in high-pressure, unpredictable environments. This study investigates the impact of team coordination and team communication on adaptive team performance. Data were collected from 230 team members using structured questionnaires and analyzed through Partial Least Squares (PLS). Results reveal significant and positive relationships between coordination and communication, and adaptive team performance. These findings highlight the importance of strengthening team functional processes to enhance responsiveness and resilience. The study provides practical insights for enhancing team effectiveness in dynamic operational environments, particularly where rapid decision-making and flexibility are essential for success in unpredictable and challenging conditions.

This paper propose and test a model of how inter-firm network effect firmpsilas absorptive capacity in a dynamic operating environment. Using data of manufacturing firms, the results indicate that the intensity of network has a positive impact on potential absorptive capacity, and the intensity of weak tie network are more important; the scale of weak tiespsila network has a positive impact on potential absorptive capacity; and the location of weak tiespsila network has a negative impact on potential absorptive capacity. The empirical results suggest that the potential absorptive capacity has mediating effects on the relationships between firm network and firmpsilas realized absorptive capacity. In addition, coordination knowledge integration mechanism has moderation effects on the relationships between firmpsilas potential absorptive capacity and realized absorptive capacity.

In the rapidly growing electronic business domain, organisations in the open environment are required to cooperate to achieve certain goals related to their business model. Given such an open environment, maintaining the privacy of organisations becomes a critical challenge. We believe the electronic marketplace (eMarketplace) is a promising architecture model for achieving cooperation between organisations. This is because of the nature of the market, where each entity selfishly attempts to achieve its own goals. The primary objective of this work is to model cooperation between different entities in the eMarketplace where the entities’ privacy requirement is inherited. The proposed architecture, the eMarketplace, exists as a collection of software agents in cooperative distributed systems. It enables and supports scheduling services between market participants. These services take into consideration any privacy desires that may be required from various participants in an open, dynamic, and heterogeneous environment.

The Business Intelligence (BI) traditional goals still stay: provide the right information users in organization with right information at the right time and place. The BI systems by their nature demonstrate as heterogeneous, while all plans to unify them eventually fail. Moreover the expectations of R&D of proactive BI let us conclude, the heterogeneity may become chaotic. We claim that, for mastering such a heterogeneous and dynamic system, a programmatically accessible and self managed knowledge management system, capable of storing and representing the whole content of a BI system is required. The properties of Knowledge Management Structures (KMS), combined with intelligent agents (IA) and Multi-agent systems (MAS) show great capabilities, and have, in limited scope, proved their strengths, acting in heterogeneous svstems. In the following paper we synthesized the state of the art research in the fields of BI, KMS, IA and tried to fill a gap disabling successful application of IA for management of KMS in a dynamical environment. To do so, a model of an IA, capable of reasoning and acting in a dynamic KMS environment and sharing its knowledge is presented.

Collision avoidance ability is very important for autonomous agricultural vehicles, but the influence of different obstacles in agricultural environment is rarely taken into account. In this paper, a velocity control strategy for collision avoidance was proposed to adjust the velocity of autonomous agricultural vehicles according to the movement state and dangerous degree of the obstacles and the distance between the obstacles and the vehicles, thus to improve intelligence and safety of the vehicles. The control strategy involved two steps: collision prediction in dynamic environments with an improved obstacle space–time grid map, and velocity generator for collision avoidance with a cloud model. Simulations were conducted on the obstacle collision prediction and the designed cloud generator for velocity control respectively. Simulation results show that the proposed strategy can effectively predict collision with anti-disturbance ability for threat-free obstacles and rapid and accurate velocity output. And it realizes the real-time operation in dynamic environments with an average time of 0.2 s to predict collision. Additionally, field experiments including five trial schemes were performed to test the proposed velocity control strategy on an agricultural robot, where a haystack, a tractor and walking persons were regarded as static or dynamic obstacles. The results of the field experiments show that the proposed velocity control strategy has strong feasibility and effectiveness.

<div>This article presents an eco-driving algorithm for electric vehicles featuring multi-speed transmissions. The proposed controller is formulated as a co-optimization problem, simultaneously optimizing both vehicle longitudinal speed and powertrain operation to maximize energy efficiency. Constraints derived from a connected vehicle–based traffic prediction algorithm are used to ensure traffic safety and smooth traffic flow in dynamic environments with multiple signalized intersections and mixed traffic. By simplifying the complex, nonlinear mixed-integer problem, the proposed controller achieves computational efficiency, enabling real-time implementation. To evaluate its performance, traffic scenarios from both Simulation of Urban MObility (SUMO) and real-world road tests are employed. The results demonstrate a notable reduction in energy consumption by up to 11.36% over an 18 km drive.</div>

Several agent frameworks have been proposed for developing intelligent software agents and multi-agent systems that are able to perform in dynamic environments. These frameworks and architectures exploit specific reasoning tasks (such as option selection, desire filtering, plan elaboration and means-end reasoning) that support agents to react, deliberate and/or interact/cooperate with other agents. Such reasoning tasks are realized by means of specific modules that agents may trigger according to circumstances, switching their behaviour between predefined discrete behavioural modes. This paper presents the facilities provided by the non-layered BDI-architecture of ICagent for supporting performance in dynamic and unpredictable multi-agent environments through efficient balancing between behavioural modes in a continuous space. This space is circumscribed by the purely (individual) reactive, the purely (individual) deliberative and the social deliberative behavioural modes. In a greater extend than existing frameworks; ICagent relates agent’s flexible behaviour to cognition and sociability, supporting the management of plans constructed by the agent’s mental and domain actions in a coordinated manner.

This paper presents theoretical and methodological issues applied to the design and analysis of didactical situations for training to manage open dynamic environment. The study deals with the case of fire-fighter officers training, who are responsible for operations against natural disasters as well as industrial or transportation accidents. Operational training aims at developing collective competencies in decision making at a tactical level in unforeseen situations. At this level, actors have to develop analytical reasoning without having direct information about the "perceived and acted" objects of action. A brief survey of the evolution of training officers at the upper level is presented. Trainees' evolution through a series of situations of simulation is then analysed through the main results of a series of questionnaires. Instructor's activity in managing situation of simulation is then analysed on the basis of the verbal interactions with trainees. Finally, the issue of designing computer-based systems as tools for instructors in managing training situations is discussed.

Communication between multiple agents is essential to achieve cooperation, negotiation, and take decisions for mutual benefit. Nowadays there is a growing interest in automating communication processes between different agents in dynamic web-based environments. However, when agents are deployed and integrated in open and dynamic environments, detailed syntax and semantics of their particular language implementations differ, causing the problem of communication heterogeneity. Therefore, it is necessary to measure heterogeneity among all participating agents and the number of required translations when heterogeneous agents are involved in communications. In this chapter we present a set of measures with the objective to evaluate the minimal computational requirements before implementing a translation approach. Our measures are based on set theory, which has proved to be a good representation formalism in other areas. We showed how to use the set of measures for two highly heterogeneous set of agents.

This research, conducted within the THETIDA project (https://thetida.eu), focuses on identifying threats posed by climate change to underwater cultural heritage sites, with a specific emphasis on the diverse risks – both direct and indirect – that endanger metallic materials, such as anchors, cannons, and structural elements. Notable pilot sites within THETIDA include the US Army WWII PB4Y-1 bomber aircraft Liberator off the coast of Praia de Faro (Portugal), the sunken submarine chaser Equa near La Spezia, and an Ottoman shipwreck in Famagusta bay, Cyprus. These sites offer varied historical backgrounds, construction materials, and biological ecosystems, enabling a comprehensive comparison between sites protected in bays and those in dynamic open ocean environments with currents, sediment dynamics, and biological actions influencing deterioration processes [1]. Being aware of successive reports of the Intergovernmental Panel on Climate Change (IPCC), documenting climate phenomena such as rising sea levels, ocean surface temperature increases, ocean acidification, changes in ocean circulation, extreme wave events and deoxygenation [2], the objectives of the research are to investigate the deterioration effects associated with climate change, assess their evolution across different environments, and develop prediction models. The ultimate goal is to provide practical recommendations for site preservation. The methodology involves studying the sites, historical backgrounds, material compositions, and deterioration characteristics, including physical, chemical, and biological factors of underwater weathering, classification of decay patterns, biocolonisation and biodeterioration characteristics through a multidisciplinary approach. In underwater heritage sites, materials undergo physical, chemical, and biological changes influenced by water, sediment, and living organisms [3, 4]. The study will consider variables such as temperature, salinity, pH, oxygen levels, and the intensity and direction of currents, alongside the depth and location of the site. Mock-up samples will undergo accelerated weathering and autoclave corrosion tests to observe the degradation related to specific environmental parameters. Mock-up samples will also be placed in the real sites to compare the results with the controlled environment aged samples and explore the early stages of deterioration and biocolonisation. This research contributes to understanding the impact of climate change on underwater cultural heritage sites, providing valuable insights for preservation efforts in the face of evolving environmental challenges. Acknowledgement: This research has been funded by European Union’s Horizon Europe research and innovation programme under the THETIDA project (Grant Agreement No. 101095253) (Technologies and methods for improved resilience and sustainable preservation of underwater and coastal cultural heritage to cope with climate change, natural hazards, and environmental pollution).  

This paper addresses the task allocation problem in open, dynamic grid or service-oriented environments. In such environments, both grid/service providers and consumers can be modeled as intelligent agents. These agents can leave off and enter into an open environment freely at any time. Task allocation under time constraints becomes a critical issue in such environments since it is difficult to apply a central controller during the allocation process due to the openness of environments as well as decentralized natures of agents. This paper proposes a negotiation-based method for task allocation under time constraints in an open, dynamic grid environment, where both consumer and provider agents can enter into or leave off the environment freely. In this method, there is no central controller, and agents negotiate with each other for task allocation based only on local views. The experimental results show that the proposed method can outperform state-of-art methods in terms of success rate of task allocation and total profit obtained from the allocated tasks by agents under different time constraints.

ABSTRACTWith the development of Internet technologies, online distance education is becoming an important format for teaching, learning, and instruction. In the 21st century, bleeding edge technologies make it possible for more and more students, especially adults, to benefit from an online education. This new paradigm in education allows students to pursue a second or third degree without sacrificing their jobs. Despite the advances in technology, existing online course management systems are far behind in the technological applications needed to deliver the highest quality online education. The existing distance education systems do not give enough consideration to many important issues, like personalization, mobility of students and instructors, and coordination among online study groups, which seriously affect further development of online education. This article discusses the potential applications of intelligent agents and other modern technologies to solve these problems and make online education more accountable in the open and dynamic environment.

When a robot manipulator performs tasks in a dynamic unstructured environment, it not only should account for where and how it should move, but also ensure that the robot can effectively avoid collisions with static and moving obstacles. Planning in absolutely unstructured and dynamic environment is much more challenging, up-to-date there is no universal solution to this problem. In this paper, we believed the concept that for motion planning in real dynamic environments, usually there is only a part of a path invalidated by obstacles, and only this part needs replanning. Thus, in this paper a practical strategy was presented for motion planning in dynamic environments. First, a variant dynamic Rapidly-exploring Random Tree (RRT) planner was designed to combine global planning with local replanning. Then, the proposed dynamic RRT algorithm was implemented under Robotic Operation System (ROS) framework and applied to a 7-degree-of-freedom manipulator. Experimental results have shown that the proposed dynamic RRT planner can plan and execute movement of robot manipulator between a starting and target position while preserving task constraints and reacting to environment changes in real time.