Tobias Warden

Towards Ontology-Based Multiagent Simulations: The Plasma Approach.

In multiagent-based simulation systems the agent programming paradigm is adopted for simulation. This simulation approach offers the promise to facilitate the design and development of complex simulations, both regarding the distinct simulation actors and the simulation environment itself. We introduce the simulation middleware PlaSMA which extends the JADE agent framework with a simulation control that ensures synchronization and provides a world model based on a formal ontological description of the respective application domain. We illustrate the benefits of an ontology grounding for simulation design and discuss further gains to be expected from recent advances in ontology engineering, namely the adaption of foundational ontologies and modelling-patterns.

SMIZE: a spontaneous ride-sharing system for individual urban transit

We introduce a ride-sharing concept for short distance travel within metropolitan areas which is designed to handle spontaneous ridesharing requests of prospective passengers with transport opportunities available on short call. The system has been designed as a multiagent system. We present a methodology to determine the feasibility of our ride-sharing approach for specific metropolitan areas and predefined operation requirements using multiagent-based simulation. Concrete experiments have been conducted for the city of Bremen, (Germany) in the FIPA-compliant multiagent-based simulation system PlaSMA.

An Agent-based Approach to Autonomous Logistic Processes Collaborative Research Centre 637: Autonomous Cooperating Logistic Processes

This paper presents the research activities of the Collaborative Research Centre (CRC) 637 “Autonomous Cooperating Logistic Processes—A Paradigm Shift and its Limitations” at the University of Bremen. After a motivation of autonomous logistics as an answer to current trends in increasingly dynamic markets, we sketch the structure and aims of the interdisciplinary CRC. We present several interpretations of the central motive of autonomous control, pursued by sub-projects over the course of the first project period, and focus on an agent-based approach to autonomous logistics.

Real-time spatio-temporal analysis of dynamic scenes

We propose a set of tools for spatio-temporal real-time analysis of dynamic scenes. It is designed to improve the grounding situation of autonomous agents in (simulated) physical domains. We introduce a knowledge processing pipeline ranging from relevance-driven compilation of a qualitative scene description to a knowledge-based detection of complex event and action sequences, conceived as a spatio-temporal pattern-matching problem. A methodology for the formalization of motion patterns and their inner composition is introduced and applied to capture human expertise about domain-specific motion situations. We present extensive experimental results from a challenging environment: 3D soccer simulation. It substantiates real-time applicability of our approach under tournament conditions, based on a 5-Hz (a) precise and (b) noisy/incomplete perception. The approach is not limited to robot soccer. Instead, it can also be applied in other fields such as experimental biology and logistic processes.

Machine learning in agent-based stochastic simulation: Inferential theory and evaluation in transportation logistics

Multiagent-based simulation is an approach to realize stochastic simulation where both the behavior of the modeled multiagent system and dynamic aspects of its environment are implemented with autonomous agents. Such simulation provides an ideal environment for intelligent agents to learn to perform their tasks before being deployed in a real-world environment. The presented research investigates theoretical and practical aspects of learning by autonomous agents within stochastic agent-based simulation. The theoretical work is based on the Inferential Theory of Learning, which describes learning processes from the perspective of a learners goal as a search through knowledge space. The theory is extended for approximate and probabilistic learning to account for the situations encountered when learning in stochastic environments. Practical aspects are exemplified by two use cases in autonomous logistics: learning predictive models for environment conditions in the future, and learning in the context of evolutionary plan optimization.

Real-Time Spatio-Temporal Analysis of Dynamic Scenes in 3D Soccer Simulation

We propose a framework for spatio-temporal real-time analysis of dynamic scenes. It is designed to improve the grounding situation of autonomous agents in (simulated) physical domains. We introduce a knowledge processing pipeline ranging from relevance-driven compilation of a qualitative scene description to a knowledge-based detection of complex event and action sequences, conceived as a spatio-temporal pattern matching problem. A methodology for the formalization of motion patterns and their inner composition is introduced and applied to capture human expertise about domain-specific motion situations. We present extensive experimental results from the 3D soccer simulation that substantiate the online applicability of our approach under tournament conditions, based on 5 Hz a) precise and b) noisy/incomplete perception.

The learnable evolution model in agent-based delivery optimization

The learnable evolution model is a stochastic optimization method which employs machine learning to guide the optimization process. LEM3, its newest implementation, combines its machine learning mode with other search operators. The presented research concerns its application within a multi-agent system for autonomous control of container on-carriage operations. Specifically, LEM3 is used by transport management agents that act on behalf of the trucks of a forwarding agency for the planning of individual transport schedules.

Agent-based Pickup and Delivery Planning: The Learnable Evolution Model Approach

The Dynamic Vehicle Routing Problem (DVRP) is an optimization problem in which agents deliver orders that are not known in advance to the routing. Partial solutions need to be adapted to continuously accommodate new orders within dynamically changing conditions. This research focuses on using a combination of multiagent-based autonomous control with non-Darwinian evolutionary optimization. In order to compile transport plans and render optimized decisions agents managing transport vehicles employ a guided evolutionary computation method, called the learnable evolution model (LEM). Implementation and experimental evaluation of the method is performed within the Plasma multiagent simulation platform.

Knowledge Management for Agent-Based Control Under Temporal Bounds

Domain-specific time limits for the execution of agent-oriented knowledge management processes constitute a significant challenge for the design of autonomous logistic control with multi-agent systems. Tailored models are needed to support the agents’ decision-making, which gives rise to questions concerning the time span agents are granted to compile these models, especially at the onset of the agent life cycle. Besides knowledge acquisition, the exploitation of the models in concrete decision situations is often subject to time limits, as well, such that efficient inference mechanisms have to be available. Finally, agents need to maintain their local models concurrently when performing logistic processes. Knowledge management tasks such as adaption of existing and compilation of new models need to be performed in a timely fashion.

Intelligent Modeling and Control for Autonomous Logistics

The increasing complexity of logistic networks calls for a paradigm change in their modeling and operations. Centralized control is no longer a feasible option when dealing with extremely large systems. For this reason, decentralized autonomous systems are gaining popularity in providing robustness and scalability. This chapter focuses on the use of intelligent systems in autonomous logistics. Specifically, it describes issues related to knowledge management, a machine learning-based approach to adaptability and planning, and intelligent optimization by autonomous logistics entities.