RoSES. Robust Self-configuring Embedded Systems ENGINEERING. Prof. Philip Koopman

Transcription

1 RoSES Robust Self-configuring Embedded Systems Prof. Philip Koopman Bill Nace Charles Shelton Meredith Beveridge Tridib Chakravarty Chris Martin Mike Bigrigg Institute for Complex Engineered Systems Electrical & Computer ENGINEERING

2 RoSES Objectives Context: distributed embedded systems Multiple smart sensors/actuators connected to embedded real-time network S A S A S A CPU CPU CPU Embedded Network CPU S A CPU S A Research goal automatic reconfiguration for: Graceful degradation Graceful reintegration after repair Graceful acceptance of non-exact spares Graceful upgrade with new capabilities 2

3 Product Family Architectures Fine grain distributed systems yield dense product lattices # Components Installed N+2 N+2 N+1 Standard Product D Product Family Standard Product B Standard Product C N N-1 Standard Product A = Product Variant = Add or Remove a Component 3

4 Why Self-Configuration? Product Families + Automatic configuration management creates a unifying capability Product families can include degradation as well as intentional price/performance tradeoff points Consider component failure as an example: Component fails triggers reconfiguration for degraded operation Component replaced reconfiguration to integrate repair part New component added reconfiguration to upgrade system That s a lot to attempt all at once Static configuration at first On-the-fly configuration as an eventual goal 4

5 A Simplistic Example Control of gasoline engine speed Complicated system controls fuel if valve is installed/operational But, baseline capability is retained in case of failure PID CONTROL ALGORITHM NETWORK CONTROLLER Degrades to BANG-BANG CONTROL ALGORITHM NETWORK FUEL VALVE IGNITION CONTROL SENSOR IGNITION CONTROL SENSOR Throttled Engine Controller Hit-or-Miss Constant Speed Engine Controller 5

6 Different Sensors / Different Capabilities Similarly, different actuators have different capabilities Mobile Object Adapters translate raw capability into desired interface LOCAL COMPUTATION, ALGORITHMS, DATA STORAGE RELATIVE SENSOR LOCAL COMPUTATION, ALGORITHMS, DATA STORAGE PROPORTIONAL SENSOR Mobile Object Adapter Mobile Object Adapters RELATIVE INTERFACE (TOO FAST/ TOO SLOW) RELATIVE INTERFACE (TOO FAST/ TOO SLOW) PERCENT MAXIMUM INTERFACE RPM INTERFACE OVER EMERGENCY INTERFACE Embedded Network Embedded Network EXAMPLE SIMPLE SENSOR EXAMPLE HIGH-END SENSOR with MULTIPLE INTERFACES 6

7 Generic RoSES System Architecture SMART SENSORS SMART ACTUATORS Basic Sensor Device Baseline Sensor SW Functionality Basic Actuator Device Baseline Actuator SW Functionality SW Adapter for High Level Logical Interface SW Adapter for High Level Logical Interface SW Adapter for High Level Logical Interface SW Adapter for High Level Logical Interface Dynamic Interface to Object Bus Dynamic Interface to Object Bus Object Bus (Run-Time Infrastructure & Network) Adapter Repository CUSTOMIZATION MANAGER Co-Scheduling & Assigment Tool 7

8 Near-Term Research Challenges Mapping functionality onto hardware Maximize utility of result given constrained resources Achieving real-time operation Co-schedule CPU, Memory, Network usage to meet real-time deadlines Achieving plug & play capabilities e.g., What would it take to put CORBA on a CAN network? Avoid re-inventing CORBA if possible Tractable demonstration Generic automotive testbed 8

9 Functionality To Hardware Mapping Automatic allocation of HW & SW components Maximize utility of functions within hardware constraints FUNCTIONAL REQUIREMENTS CONTROL ALGORITHMS & SOFTWARE Dynamic Adaption Runtime System CPUs MEMORY NETWORK SENSORS ACTUATORS 9

10 Proposed Testbed Navigation + active vehicle stability control Inertial sensors / dead reckoning subsystem 3d inertial platform with acceleration and speed readouts Steering angle Wheel speed GPS-based navigation subsystem External source of position and speed CANalyzer Test Tool Data collection subsystem Inertial Navigation Stores and forwards data for failure diagnosis Gateway to wireless internet connection Simulated using Wireless Andrew CAN Embedded Network Precision Navigation Synthesis Data Collection Vehicle Stability Control CANalyzer system to simulate rest of vehicle network Provides messages for realistic operating environment GPS Navigation Wireless Andrew 10

11 Experiments Applications Phase 1: Precision navigation Gracefully degrading navigation based on sensor information Combine inertial & GPS for result better than either alone Phase 2: Active vehicle stability control Vehicle stability algorithms vary degree of control based on quality of sensor information Graceful degradation rather than brute force redundancy Year 1 goal lab demo: Demonstrate automatic reconfiguration when sensors/actuators/ computers fail for navigation application Later goals: Demonstrate automatic reconfiguration in a test vehicle Demonstrate upgrade of baseline component with advanced/proprietary component Demonstrate graceful failure and restoration after repair in a test vehicle 11

12 Possible Extensions Embedded system point of view for system security Vehicle / external firewall Protection against malicious or faulty components on networks Safe upgrade strategies Even if new components have software defects (?!) 12

13 People Prof. Phil Koopman 10+ years industry experience embedded hardware & software Bill Nace Ph.D. student: functionality-to-resource mapping Charles Shelton Ph.D. student: software architecture Meredith Beveridge M.S. student: testbed/baseline applications + Jini-on-CAN Chris Martin M.S. student: simulation infrastructure (details TBD) Tridib Chakravarty M.S. student: embedded protocol infrastructure Mike Bigrigg staff member R o S E S 13