Real-Time BDI Agents: A Model and Its Implementation
Authors: Andrea Traldi, Francesco Bruschetti, Marco Robol, Marco Roveri, Paolo Giorgini
IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We propose an implementation of the model for a resource-collection video-game and we validate the approach against a set of significant scenarios. To validate our proposed solution we deployed it in a video-game setting, where each agent in the game executes on top of a unique real-time execution environment with limited computation capabilities, that is possibly running other real-time tasks. |
| Researcher Affiliation | Academia | Andrea Traldi1 , Francesco Bruschetti1 , Marco Robol2 , Marco Roveri2 and Paolo Giorgini2 University of Trento 1{andrea.traldi, francesco.bruschetti}@studenti.unitn.it, 2{marco.robol, marco.roveri, paolo.giorgini}@unitn.it |
| Pseudocode | Yes | Algorithm 1 RT-BDI Reasoning cycle |
| Open Source Code | Yes | The extended version of this paper and all the code are available at https://rti-bdi.github.io/ijcai2022/. Code available at https://github.com/RTI-BDI/AT-Kronity. |
| Open Datasets | No | The paper validates its approach using a custom-built resource-collection video game called Kronity, described as a 'synthetic scenario,' but does not provide access information or citations for a publicly available dataset. |
| Dataset Splits | No | The paper uses the term 'validation' in the context of verifying its approach against scenarios, not to describe dataset splits. It does not provide specific data split information (percentages, sample counts, or methodology) for training, validation, or testing datasets. |
| Hardware Specification | No | The paper mentions 'a unique real-time execution environment with limited computation capabilities' for its experiments but does not provide any specific hardware details such as CPU/GPU models, memory, or cloud instance types. |
| Software Dependencies | No | The paper mentions software components like 'Unity', 'Kronosim', 'PDDL 2.1', and 'OPTIC', but it does not provide specific version numbers for these software dependencies (e.g., Unity version, Kronosim version, OPTIC version). |
| Experiment Setup | Yes | Section 3, 'The Real-Time BDI Architecture', describes the BDI layer's components like desires, goals, and intentions (including their properties like deadlines and priorities), and how the BDI Reasoning Loop operates (Algorithm 1). Section 4, 'The Real-Time BDI Agents Simulator', details how Kronosim was extended and deployed within Unity, and how it interacts with the environment, including the invocation of the OPTIC planner and the conversion of temporal plans. Section 5, 'Validating the RT-BDI Agent Architecture', further describes specific scenarios and their execution logs, outlining how the agents are configured to interact within the simulation. |