Francesca Cairoli
Nicola Paoletti
ABSTRACT
We consider the problem of predictive monitoring (PM), i.e., predicting at runtime the satisfaction of a desired property from the current system’s state. Due to its relevance for runtime safety assurance and online control, PM methods need to be efficient to enable timely interventions against predicted violations, while providing correctness guarantees. We introduce quantitative predictive monitoring (QPM), the first PM method to support stochastic processes and rich specifications given in Signal Temporal Logic (STL). Unlike most of the existing PM techniques that predict whether or not some property ϕ is satisfied, QPM provides a quantitative measure of satisfaction by predicting the quantitative (aka robust) STL semantics of ϕ. QPM derives prediction intervals that are highly efficient to compute and with probabilistic guarantees, in that the intervals cover with arbitrary probability the STL robustness values relative to the stochastic evolution of the system. To do so, we take a machine-learning approach and leverage recent advances in conformal inference for quantile regression, thereby avoiding expensive Monte Carlo simulations at runtime to estimate the intervals. We also show how our monitors can be combined in a compositional manner to handle composite formulas, without retraining the predictors or sacrificing the guarantees. We demonstrate the effectiveness and scalability of QPM over a benchmark of four discrete-time stochastic processes with varying degrees of complexity.
- Alessandro Abate, Henk Blom, Marc Bouissou, Nathalie Cauchi, Hassane Chraibi, Joanna Delicaris, Sofie Haesaert, Arnd Hartmanns, Mahmoud Khaled, Abolfazl Lavaei, 2021. ARCH-COMP21 Category Report: Stochastic Models. In 8th International Workshop on Applied Verification of Continuous and Hybrid Systems, ARCH 2021. EasyChair, 55–89.Google Scholar
- Alessandro Abate, Joost-Pieter Katoen, John Lygeros, and Maria Prandini. 2010. Approximate model checking of stochastic hybrid systems. European Journal of Control 16, 6 (2010), 624–641.Google ScholarCross Ref
- Vineeth Balasubramanian, Shen-Shyang Ho, and Vladimir Vovk. 2014. Conformal prediction for reliable machine learning: theory, adaptations and applications. Newnes.Google Scholar
- Ezio Bartocci, Luca Bortolussi, Laura Nenzi, and Guido Sanguinetti. 2015. System design of stochastic models using robustness of temporal properties. Theoretical Computer Science 587 (2015), 3–25.Google ScholarDigital Library
- Ezio Bartocci, Jyotirmoy Deshmukh, Alexandre Donzé, Georgios Fainekos, Oded Maler, Dejan Ničković, and Sriram Sankaranarayanan. 2018. Specification-based monitoring of cyber-physical systems: a survey on theory, tools and applications. In Lectures on Runtime Verification. Springer, 135–175.Google Scholar
- Luca Bortolussi, Francesca Cairoli, Ginevra Carbone, and Paolo Pulcini. 2022. Stochastic Variational Smoothed Model Checking. arXiv preprint arXiv:2205.05398 (2022).Google Scholar
- Luca Bortolussi, Francesca Cairoli, Nicola Paoletti, Scott A Smolka, and Scott D Stoller. 2019. Neural Predictive Monitoring. In International Conference on Runtime Verification. Springer, 129–147.Google Scholar
- Luca Bortolussi, Francesca Cairoli, Nicola Paoletti, Scott A Smolka, and Scott D Stoller. 2021. Neural predictive monitoring and a comparison of frequentist and Bayesian approaches. International Journal on Software Tools for Technology Transfer 23, 4 (2021), 615–640.Google ScholarDigital Library
- Luca Bortolussi, Dimitrios Milios, and Guido Sanguinetti. 2016. Smoothed model checking for uncertain continuous-time Markov chains. Information and Computation 247 (2016), 235–253.Google ScholarDigital Library
- Francesca Cairoli, Luca Bortolussi, and Nicola Paoletti. 2021. Neural Predictive Monitoring under Partial Observability. arxiv:2108.07134 [cs.LG]Google Scholar
- Francesca Cairoli, Nicola Paoletti, and Luca Bortolussi. 2022. Conformal Quantitative Predictive Monitoring of STL Requirements for Stochastic Processes. arXiv preprint arXiv:2211.02375 (2022).Google Scholar
- Francesca Cairoli, Nicola Paoletti, and Luca Bortolussi. 2022. Neural Predictive Monitoring for Collective Adaptive Systems. In ISoLA 2022 Symposium, to appear.Google Scholar
- Yi Chou, Hansol Yoon, and Sriram Sankaranarayanan. 2020. Predictive Runtime Monitoring of Vehicle Models Using Bayesian Estimation and Reachability Analysis. In Intl. Conference on Intelligent Robots and Systems (IROS).Google Scholar
- Badis Djeridane and John Lygeros. 2006. Neural approximation of PDE solutions: An application to reachability computations. In Proceedings of the 45th IEEE Conference on Decision and Control. IEEE, 3034–3039.Google ScholarCross Ref
- Alexandre Donzé and Oded Maler. 2010. Robust Satisfaction of Temporal Logic over Real-Valued Signals. In Proceedings of International Conference on Formal Modeling and Analysis of Timed Systems. Springer, Klosterneuburg, Austria, 92–106. https://doi.org/10.1007/978-3-642-15297-9_9Google ScholarCross Ref
- Ansgar Fehnker and Franjo Ivančić. 2004. Benchmarks for hybrid systems verification. In International Workshop on Hybrid Systems: Computation and Control. Springer, 326–341.Google ScholarCross Ref
- Wolfgang Granig, Stefan Jakšić, Horst Lewitschnig, Cristinel Mateis, and Dejan Ničković. 2020. Weakness Monitors for Fail-Aware Systems. In International Conference on Formal Modeling and Analysis of Timed Systems. Springer, 283–299.Google Scholar
- Radoslav Ivanov, James Weimer, Rajeev Alur, George J Pappas, and Insup Lee. 2019. Verisig: verifying safety properties of hybrid systems with neural network controllers. In Proceedings of the 22nd ACM International Conference on Hybrid Systems: Computation and Control. 169–178.Google ScholarDigital Library
- Volodymyr Kuleshov, Nathan Fenner, and Stefano Ermon. 2018. Accurate uncertainties for deep learning using calibrated regression. In International Conference on Machine Learning. PMLR, 2796–2804.Google Scholar
- Pavel Kurasov, Alexander Lück, Delio Mugnolo, and Verena Wolf. 2018. Stochastic hybrid models of gene regulatory networks–a pde approach. Mathematical biosciences 305 (2018), 170–177.Google Scholar
- Lars Lindemann, Alena Rodionova, and George Pappas. 2022. Temporal Robustness of Stochastic Signals. In 25th ACM International Conference on Hybrid Systems: Computation and Control. 1–11.Google Scholar
- Meiyi Ma, John Stankovic, Ezio Bartocci, and Lu Feng. 2021. Predictive monitoring with logic-calibrated uncertainty for cyber-physical systems. ACM Transactions on Embedded Computing Systems (TECS) 20, 5s (2021), 1–25.Google ScholarDigital Library
- Anirudha Majumdar and Marco Pavone. 2020. How should a robot assess risk? towards an axiomatic theory of risk in robotics. In Robotics Research. Springer, 75–84.Google Scholar
- Oded Maler and Dejan Nickovic. 2004. Monitoring temporal properties of continuous signals. In Formal Techniques, Modelling and Analysis of Timed and Fault-Tolerant Systems. Springer, 152–166.Google Scholar
- Roland Malhame and Chee-Yee Chong. 1985. Electric load model synthesis by diffusion approximation of a high-order hybrid-state stochastic system. IEEE Trans. Automat. Control 30, 9 (1985), 854–860.Google ScholarCross Ref
- Harris Papadopoulos. 2008. Inductive conformal prediction: Theory and application to neural networks. In Tools in artificial intelligence. InTech.Google Scholar
- Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, 2019. PyTorch: An imperative style, high-performance deep learning library. In Advances in Neural Information Processing Systems. 8024–8035.Google Scholar
- Dung Phan, Nicola Paoletti, Timothy Zhang, Radu Grosu, Scott A Smolka, and Scott D Stoller. 2018. Neural state classification for hybrid systems. In International Symposium on Automated Technology for Verification and Analysis. Springer, 422–440.Google ScholarCross Ref
- Dung T Phan, Radu Grosu, Nils Jansen, Nicola Paoletti, Scott A Smolka, and Scott D Stoller. 2020. Neural simplex architecture. In NASA Formal Methods Symposium. Springer, 97–114.Google ScholarDigital Library
- Xin Qin and Jyotirmoy V Deshmukh. 2019. Predictive monitoring for signal temporal logic with probabilistic guarantees. In Proceedings of the 22nd ACM International Conference on Hybrid Systems: Computation and Control. ACM, 266–267.Google ScholarDigital Library
- Xin Qin and Jyotirmoy V Deshmukh. 2020. Clairvoyant Monitoring for Signal Temporal Logic. In International Conference on Formal Modeling and Analysis of Timed Systems. Springer, 178–195.Google Scholar
- Vasumathi Raman, Alexandre Donzé, Mehdi Maasoumy, Richard M. Murray, Alberto Sangiovanni-Vincentelli, and Sanjit A. Seshia. 2014. Model predictive control with signal temporal logic specifications. In Proceedings of the IEEE Conference on Decision and Control. IEEE, Los Angeles, USA, 81–87. https://doi.org/10.1109/CDC.2014.7039363Google ScholarCross Ref
- Alëna Rodionova, Lars Lindemann, Manfred Morari, and George J Pappas. 2021. Time-robust control for STL specifications. In 2021 60th IEEE Conference on Decision and Control (CDC). IEEE, 572–579.Google ScholarDigital Library
- Yaniv Romano, Evan Patterson, and Emmanuel J Candès. 2019. Conformalized quantile regression. arXiv preprint arXiv:1905.03222 (2019).Google Scholar
- Vicenc Rubies Royo, David Fridovich-Keil, Sylvia Herbert, and Claire J Tomlin. 2018. Classification-based Approximate Reachability with Guarantees Applied to Safe Trajectory Tracking. arXiv preprint arXiv:1803.03237 (2018).Google Scholar
- Vladimir Vovk, Alex Gammerman, and Glenn Shafer. 2005. Algorithmic learning in a random world. Springer Science & Business Media.Google Scholar
- E. Yel, T. J. Carpenter, C. Di Franco, R. Ivanov, Y. Kantaros, I. Lee, J. Weimer, and N. Bezzo. 2020. Assured Runtime Monitoring and Planning: Toward Verification of Neural Networks for Safe Autonomous Operations. IEEE Robotics & Automation Magazine 27, 2 (2020), 102–116.Google ScholarCross Ref
- Hansol Yoon and Sriram Sankaranarayanan. 2021. Predictive runtime monitoring for mobile robots using logic-based bayesian intent inference. In 2021 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 8565–8571.Google ScholarDigital Library
- Håkan LS Younes, Marta Kwiatkowska, Gethin Norman, and David Parker. 2006. Numerical vs. statistical probabilistic model checking. International Journal on Software Tools for Technology Transfer 8, 3 (2006), 216–228.Google ScholarDigital Library
- Xinyi Yu, Weijie Dong, Xiang Yin, and Shaoyuan Li. 2022. Model Predictive Monitoring of Dynamic Systems for Signal Temporal Logic Specifications. arXiv preprint arXiv:2209.12493 (2022).Google Scholar
Recommendations
Conformal Prediction for STL Runtime Verification
ICCPS '23: Proceedings of the ACM/IEEE 14th International Conference on Cyber-Physical Systems (with CPS-IoT Week 2023)We are interested in predicting failures of cyber-physical systems during their operation. Particularly, we consider stochastic systems and signal temporal logic specifications, and we want to calculate the probability that the current system ...
Outcome-Oriented Predictive Process Monitoring: Review and Benchmark
Predictive business process monitoring refers to the act of making predictions about the future state of ongoing cases of a business process, based on their incomplete execution traces and logs of historical (completed) traces. Motivated by the ...
Predictive monitoring environment
CSCC'11: Proceedings of the 2nd international conference on Circuits, Systems, Communications & ComputersPredictive monitoring is a technology that helps prevent/favor the occurrence of undesired/desired events: a subject (human being, machine, picture, etc.) is regularly monitored and a suitable system provides predictions about the possible future ...
Comments