A domain-specific software architecture for a class of intelligent patient monitoring agents

Abstract

A good software architecture facilitates application system development, promotes achievement of functional requirements, and supports system reconfiguration. We present a domain-specific software architecture (DSSA) that we have developed for a large application domain of adaptive intelligent systems (AISs). The DSSA provides: (a) an AIS reference architecture designed to meet the functional requirements shared by applications in this domain, (b) principles for decomposing expertise into highly reusable components, and (c) an application configuration method for selecting relevant components from a library and automatically configuring instances of those components in an instance of the architecture. The AIS reference architecture incorporates features of layered, pipe and filter, and blackboard style architectures. We describe three studies demonstrating the utility of our architecture in the sub-domain of mobile office robots and identify software engineering principles embodied in the architecture.

40 years of cognitive architectures: core cognitive abilities and practical applications

“…Activation mechanisms contribute to modeling many properties of working memory, such as limited capacity, temporal decay, rapid updating as the circumstances change, connection to other memory components and decision-making. Another common paradigm, the blackboard architecture, represents memory as a shared repository of goals, problems and partial results which can be accessed and modified by the modules running in parallel (AIS [213], CERA-CRANIUM [23], CoSy [490], FORR [137], Ymir [536], LIDA [171], ARCADIA [46], Copycat/Metacat [350], CHARISMA [102], PolyScheme [94], PRS [183]). The solution to the problem is obtained by continuously updating the shared short-term storage with information from specialized heterogeneous modules analogous to a group of people completing a jigsaw puzzle.…”

The CORTEX cognitive robotics architecture: Use cases

“…The paradigm has been modified for working well in situations where many sources of information must be combined to solve complex real-time tasks. In the AIS architecture [14], two levels are differentiated: a cognitive and a physical level. The blackboard scheme is used as an implementation of the cognitive layer with a central working memory, where all behaviours (methods) write (perceptual inputs or previously executed behaviours) and to which all behaviours are connected to be triggered when certain conditions are fulfilled.…”

Agents in a nutshell-a very brief introduction

“…Agents can help in these types of tasks because they do not get bored when nothing is happening, and during crises they can help to manage information overload. Examples of monitoring agents include Hayes-Roth and Larsson's agent [16], which monitors patients in intensive care units, and ASPIRE (created by Jones and Jacobs [20]), an agent that can be delegated to monitor satellite operations. The degree to which people are willing to delegate duties to monitoring agents depends on how much they trust the capabilities of that agent.…”

“…Early architectures were created by Brooks [5] and Kaelbling [21]. More recent examples include LOOM [25], a language for modeling agents; Hayes-Roth and Larsson's domain-specific agent architecture [16] for medical monitoring tasks; Harandi and Rendon's intermediate level architecture [15], which provides rich, yet general-purpose, support; and Sloman's global architecture [37] to enable construction of agents with human-like capabilities.…”

“…Given that agents may require a lot of work to build, there is a strong incentive to look for ways to reuse parts of existing agent systems in other systems. Sycara and Decker examines reusable behaviors for agents [35], and Hayes-Roth and Larsson [16] examines reconfigurable components of expertise.…”