作者: Nick Hawes , Jeremy L. Wyatt , Mohan Sridharan , Henrik Jacobsson , Richard Dearden
DOI: 10.1007/978-3-642-11694-0_2
关键词: Robot architecture 、 Architecture 、 Human–computer interaction 、 Intelligent decision support system 、 Systems engineering 、 Design space 、 System building 、 Computer science 、 Binding problem 、 Partially observable Markov decision process 、 Space (commercial competition)
摘要: The study of architectures to support intelligent behaviour is certainly the broadest, and arguably one most ill-defined enterprises in AI Cognitive Science. basic scientific question we seek answer is: “What are trade-offs between different ways that systems might be structured?” These depend large part on what kinds tasks environment a system operates under (niche space), also aspects design space deem architectural. In CoSy have tried several ways. First, by thinking about requirements arise from our particular scenarios (parts niche space). Second, building follow well-defined architectural rules, using these carry out experiments variations those rules. Third, insights improve understanding choices, i.e. partial designs. Our objective has not been come up with just another robot architecture, but instead try make some small steps forward new science architectures.