Paper reading and Critiques : The Architecture Analysis & design language (AADL): An introduction

The Architecture Analysis & design language (AADL): An introduction

AADL is designed for the specification, analysis, and automated integration of real-time performance-critical (timing, safety, schedulability, fault tolerant, security, etc.) distributed computer systems. It allows analysis of system designs (and system of systems) prior to development. Also, it supports a model-based, model-driven development approach throughout the system life cycle.

• The paper lacks reasoning for providing support for embedded real time system, i.e., if a reader has no knowledge about the initial “need” (history) of having a formal DL for embedded systems, he might not fully understand existence of non-functional support by the AADL like schedulability, performance etc. and so might not be able to fully exploit the system for his system’s modal. 
• Eventually when I searched over internet I found that the basic motivation behind AADL is modeling electronics’ time-critical performance oriented system in which the software is tightly bound to its hardware.
• In terms of mapping the software model with the execution counterpart, there is no relevant information so again one has to dig deep into AADL supporting websites.
• Authors described basic support and features for AADL in textual and graphical forms with relevant examples, which help us to relate usage with significance and presentation. For example, significance of annex, components, process, flows, calls, connections etc.  Help an architect decide when and under what circumstances should he incorporate those features.
• AADL supports MetaH’s features like extensibility exposed via annex, dynamic support via mode switching, non-functional properties of both hardware and software counterparts via features like schedulability (via periodic/aperiodic etc.). This is the one of the first architectures supporting real time critical embedded systems. Even Boeing is using the same for its product line. I hope to read more papers on AADL and perhaps architectures of some real systems build on it.