Architecture-Driven Self-Adaptation and Self-Management in Robotics Systems By George Edwards, Joshua Garcia, Farshad Tajalli, Daniel Popescu, Nenad Medvidovic, Gaurav Sukhatme, Brad Petrus Presentation by Joshua Garcia and Farshad Tajalli
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Architecture-Driven Self-Adaptation and Self-Management in Robotics Systems
By George Edwards, Joshua Garcia, Farshad Tajalli,Daniel Popescu, Nenad Medvidovic, Gaurav Sukhatme, Brad Petrus
Presentation by Joshua Garcia and Farshad Tajalli
Today’s Software Systems
• 24/7 systems
• Long-lived, decentralized, heterogeneous, mobile,
ubiquitous
2
Self-awareness and Self-adaptation
Software Architecture for Robotics
3
Self-awareness and Self-adaptation
D. Kim et al. SHAGE: a framework for self-managed robot software. In Proceedings of SEAMS 2006, pages 79–85. ACM New York, NY, USA, 2006.
J. Georgas and R. Taylor. Policy-based self-adaptive architectures: a feasibility study in the robotics domain. In Proceedings of SEAMS 2008, pages 105–112. ACM New York, NY, USA, 2008.
D. Sykes et al. From goals to components: a combined approach to self-management. In Proceedings of SEAMS 2008, pages 1–8. ACM New York, NY, USA, 2008.
Layered Architecture
5
Approach
o Layered Architecture
o Component-based Approach
o Meta-level Components vs. Application Components
o Meta-level Component Types
• Collector, Analyzer, Admin
o Meta-level Component Layering
o Meta-level components form a control loop on the lower level architecture
o Implementation Support for Meta-level component types and layers
6
Approach Advantages
o Architectural Awareness
• Architectural awareness of components in the layer below
o Self-adaptation support for meta-level component types
o Self-adaptation support for qualities of service
o Support for Adaptive Reference Architectures
7
Meta-Level Components
Meta-Level Components Types
o Collector
o Analyzer
o Admin
8
Collector
o Read access to lower architecture
o A collection of Monitors
• added to the Components and Connectors in the lower layer
• exceptions, variables and events
o Collector aggregates the monitored data
o Flexibility of monitor installation
9
Analyzer
o Read access to lower architecture
o A collection of Policies
o Policies are evaluated against monitored data
received from Collectors
o Triggers adaptations and reconfigurations
o Directs Admins to modify the lower architecture
10
Admin
o Read\Write access to lower architecture
o Adapts the lower architecture
• Add, delete, connect, disconnect
o Reconfigures the lower architecture
components, connectors and etc.
11
Layered Architecture
12
Fault Tolerance Example
13
IComponentIConnector
Scaffold
AbstractDispatcher
Round RobinDispatcher
AbstractScheduler
FifoScheduler
Brick
Architecture
ExtensibleComponent
Component
Connector
Event
Port
IPort
Serializable
IArchitecture
#mutualPort
Implementation Support – Prism-MW
16
• Meta-level component type adaptation• Robotics-specific meta-level component types Meta-level
Component
Three Layer Reference Model
19
Application to Three Layer Reference Model
20
Application Scenario
21
Conclusion
o Meta-level component types and layering
• Support for qualities of service
• Ensures correct functionality and system stability
• Implementation support for meta-level component
types and layers
• Support for adaptive reference architectures
22
Related Documents
