Runtime Verification - 6th International Conference, RV 2015 Vienna, Austria, September 22-25, 2015. Proceedings

Ezio Bartocci, R MajumdarRv 2015 · DOI · DBLP

Introduction

The 6th International Conference on Runtime Verification (RV 2015), held in Vienna, Austria, serves as a cornerstone for the development of "lightweight" formal methods. Unlike traditional static verification, which analyzes code before execution, runtime verification (RV) focuses on monitoring systems as they run. This volume of proceedings captures a significant shift in the field toward practical, real-time applications in safety-critical domains.

The conference was led by chairs Ezio Bartocci (TU Wien) and Rupak Majumdar (Max Planck Institute).

The Limitations of Static Analysis

Traditional formal methods, such as model checking and theorem proving, are powerful but often struggle with the "state space explosion" found in industrial-scale software. Moreover, they cannot easily account for the unpredictable environmental factors that affect cyber-physical systems, such as autonomous vehicles or medical devices.

RV addresses this by providing a mathematical safety net. It answers critical questions: How can we monitor behavior without significant performance lag? Can we use formal logic to steer a system back to safety? And how do we handle "Heisenbugs"—errors that disappear when you attempt to observe them?

Methodological Innovations

The proceedings detail several cutting-edge techniques for system monitoring:

  • Temporal Logics (MTL and STL): Metric and Signal Temporal Logics are used to define requirements for continuous signals, essential for monitoring glucose levels in artificial pancreas controllers or sensor data in robotics.
  • Parametric Trace Slicing: This method breaks down massive execution logs into manageable "slices" based on specific parameters, allowing for efficient, real-time analysis.
  • Statistical Model Checking: By estimating the probability of property satisfaction, this approach offers a faster alternative to exhaustive verification.
  • Causality Analysis: A hybrid technique that identifies the root cause of failures by analyzing the chain of events leading to a runtime violation.

Real-World Impact and Benchmarks

A major theme of RV 2015 is the application of these methods to life-critical systems. For instance, research on the "Artificial Pancreas" demonstrates how RV can prevent fatal errors in insulin delivery.

The volume also highlights the maturity of the field through standardized benchmarks like CRV 2015 (the 2nd International Competition on Runtime Verification) and the RERS Challenge, which provide complex, synthesized systems for rigorous testing.

Key Results and Tool Maturity

The conference maintained high standards, accepting only 33% of regular paper submissions. This selectivity ensures that the introduced tools, such as S-Taliro and STARVOORS, are robust enough for production use.

The RV 2015 Program Committee included experts from prestigious global institutions, reflecting the international collaboration in the field.

Crucially, several papers demonstrated that formal monitoring can be achieved with minimal overhead—often less than 5-10% CPU impact—making it viable for high-performance environments.

Future Directions

While RV has made great strides, the "observer overhead" remains a challenge for high-frequency systems. Future work is expected to focus on predictive monitoring, moving from detecting violations after they happen to foreseeing them before they occur. Additionally, monitoring distributed, asynchronous systems without a global clock remains a key area for ongoing research.

Summary updated

BibTeX
@inproceedings{bartocci2015runtime,
  author = {Ezio Bartocci and R Majumdar},
  title = {{Runtime Verification - 6th International Conference, RV 2015 Vienna, Austria, September 22-25, 2015. Proceedings}},
  booktitle = {Rv},
  year = {2015},
  doi = {10.1007/978-3-319-23820-3},
}
Figure