Stream-Based Perception for Agents on Mobile Devices
| Autor: | Ralf Bruns, Jürgen Dunkel, Jeremias Dötterl, Sascha Ossowski |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Computer science
Multi-agent system media_common.quotation_subject Mobile computing Complex system 020206 networking & telecommunications 02 engineering and technology Deliberation Bridge (nautical) Human–computer interaction Perception 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Abstraction Mobile device media_common |
| Zdroj: | Agreement Technologies ISBN: 9783030172930 AT |
| DOI: | 10.1007/978-3-030-17294-7_6 |
| Popis: | Multi-Agent Systems (MAS) lack advanced concepts for data stream processing, which inhibits their effective use in mobile ecosystems, where built-in smartphone sensors can provide valuable data about the current physical environment of the mobile user. With beliefs, plans, and goals, cognitive agent frameworks provide useful abstractions for the development of complex systems but do not contain effective mechanisms to sufficiently bridge the abstraction gap that exists between low-level streaming data and high-level percepts. The main contribution of this paper is an enhanced perception approach, which integrates two new abstractions, namely expectations and interpretations, into the commonly used perceive-deliberate-act cycle. Expectations and interpretations address the challenges of sensor data and provide higher-level knowledge to the agent’s deliberation, which allows mobile agents to make situation-aware decisions in dynamically changing environments. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|