The Rare Glitch Project Verification Tools for Embedded

The Rare Glitch Project: Verification Tools for Embedded Systems Ed Clarke, David Garlan, Bruce Krogh, Reid Simmons, Jeannette Wing Carnegie Mellon University Pittsburgh, PA

Embedded Systems l They are highly distributed, multi-task, concurrent real-time systems. l They control increasingly sophisticated physical systems. l They operate with increasing autonomy under adverse and unpredictable conditions. l They eventually will be seamlessly woven into the fabric of our everyday lives. Carnegie Mellon: The Rare Glitch Project 2 Jeannette M. Wing

Verification l Grand Challenge ¤ To ensure the correctness of these embedded systems as they operate in increasingly complex environments—for the safety and good of the general public. l Opportunity ¤ State of the art verification tools are well suited for control systems. ¤ Embedded systems have narrow hardware/software interface. ¤ Embedded systems architectures are simpler and more regular. l Feasibility ¤ We are not trying to solve the whole verification problem, but one for a class of systems for which we believe we can make the next big success story. Carnegie Mellon: The Rare Glitch Project 3 Jeannette M. Wing

Our Long-Term Vision l To provide design engineers with lightweight “pushbutton” tools, each checking a specific applicationspecific property. Check Restart Deadlock Race Carnegie Mellon: The Rare Glitch Project Check 4 Check Power usage Fuel usage Jeannette M. Wing

Three Main Thrusts of Our Project system environment l Verifying system integrity Synchronization constraints ¤ Resource constraints ¤ Real-time constraints ¤ l Modeling the environment Hybrid dynamics ¤ Stochastic models ¤ l Usability Extracting models ¤ Explaining tool feedback ¤ Carnegie Mellon: The Rare Glitch Project 5 Jeannette M. Wing

Technical Challenges l Interleavings of multiple task executions l Inter-task dependencies and synchronization requirements l System resource constraints l Hard real-time constraints l Interactions with complex physical dynamic systems l Adverse, unpredictable environments l Stringent requirements for autonomy, fault tolerance, and survivability Carnegie Mellon: The Rare Glitch Project 6 Jeannette M. Wing

Technical Approach Embedded System Environment Design Engineer’s Verification Toolbox Usability System Integrity Properties Synchronization Constraints Resource Constraints Modeling the Environment System-Environment Modeling Uncertainty: Interactions: Hybrid Stochastic Models Dynamics Real-time Constraints Foundational Building Blocks Model checking techniques and tools Task-level system architecture, task executive languages Continuous dynamics, quantitative analysis, reliability analysis Carnegie Mellon: The Rare Glitch Project 7 Jeannette M. Wing

Cross-Cutting Themes l Exploit model checking techniques and tools. l Focus on the task level. l Incorporate continuous domains in our models. Carnegie Mellon: The Rare Glitch Project 8 Jeannette M. Wing

Verifying System Integrity l Synchronization constraints ¤ Publish-subscribe architecture ¤ Cyclic-tasks with shared variables ¢ Research question: How can we reason about their aggregate behavior? l Resource constraints ¤ Continuous, consumable (e. g. , fuel) and renewable (e. g. , bandwidth, disk space) resources ¢ Research question: How should we specify and verify resource constraints? l Real-time constraints ¤ Quantitative ¢ timing analysis Research questions: How can we extend this technique to continuous time? How can we deal with dynamic process creation and deletion? Carnegie Mellon: The Rare Glitch Project 9 Jeannette M. Wing

Modeling the Environment l Hybrid dynamics ¤ Dealing with both discrete and continuous state variables ¢ Research question: How can we automate and scale differential equations models for large systems? l Stochastic models ¤ Dealing with uncertainty, unpredictability ¤ Using continuous variables (e. g. , probabilities) for reliability and cost-benefit analyses ¢ Research questions: How can we handle dependent events, cascading of events? Both nondeterministic and probabilistic state transitions? Carnegie Mellon: The Rare Glitch Project 10 Jeannette M. Wing

Usability l Extracting models ¤ Synchronization skeletons ¤ Translators for task execution languages ¢ Research question: How can we balance expressibility of languages with efficiency of verification and user-friendliness? l Explaining verification tool results ¤ Adopt AI techniques to explain counterexamples for task executive programs. ¤ Develop techniques to view and browse the temporal evolution of counterexamples. ¢ Research question: How can we create virtual execution traces that can be used by existing visualization tools? Carnegie Mellon: The Rare Glitch Project 11 Jeannette M. Wing

The Rare Glitch Tool Suite Checkers and Provers Specification and Modeling Languages Specification Analysis Engines Checkmate Reliability and Cost Analyzers Prism SMV Counterexamples Model Explanation Generator PVS Abstraction/ Refinement Processor Symp … … Carnegie Mellon: The Rare Glitch Project 12 Jeannette M. Wing

Preliminary Planning l Technical ¤ So far: Technical presentations of our current work ¤ To do ¢ ¢ Identification of case studies Identification of common group project ¤ Continue to pursue existing two-way and three-way collaborations l Administrative ¤ Weekly page project meetings, project name, project web Carnegie Mellon: The Rare Glitch Project 13 Jeannette M. Wing

Project Administration l Principal Investigators ¤ Ed Clarke, David Garlan, Bruce Krogh, Reid Simmons, Jeannette Wing (POC) l Industrial connections ¤ Honeywell Technology Center ¤ CMU High Dependability Computing Consortium l Years 3 and 5 ¤ Hold workshops for technical exchange with industry Carnegie Mellon: The Rare Glitch Project 14 Jeannette M. Wing

Preview of This Afternoon’s Talks l Ed Clarke ¤ Verifying Systems Bus Protocol Standards for Embedded l Bruce Krogh ¤ Model Checking for Hybrid Systems l Jeannette Wing ¤ Scenario Graph Generation and MDP-Based Analysis l David Garlan ¤ Model Checking Publish-Subscribe Software Architectures Carnegie Mellon: The Rare Glitch Project 15 Jeannette M. Wing