Predictive Monitoring with Logic-Calibrated Uncertainty for Cyber-Physical Systems

  title={Predictive Monitoring with Logic-Calibrated Uncertainty for Cyber-Physical Systems},
  author={Meiyi Ma and John A. Stankovic and Ezio Bartocci and Lu Feng},
  journal={ACM Transactions on Embedded Computing Systems (TECS)},
  pages={1 - 25}
  • Meiyi Ma, J. Stankovic, Lu Feng
  • Published 31 October 2020
  • Computer Science
  • ACM Transactions on Embedded Computing Systems (TECS)
Predictive monitoring—making predictions about future states and monitoring if the predicted states satisfy requirements—offers a promising paradigm in supporting the decision making of Cyber-Physical Systems (CPS). Existing works of predictive monitoring mostly focus on monitoring individual predictions rather than sequential predictions. We develop a novel approach for monitoring sequential predictions generated from Bayesian Recurrent Neural Networks (RNNs) that can capture the inherent… 
Neural Predictive Monitoring under Partial Observability
This work presents a learning-based method for PM that produces accurate and reliable reachability predictions despite partial observability (PO), and uses the latter to identify unreliable (and likely erroneous) predictions and to retrain and improve the monitors on these uncertain inputs.
Systems Challenges for Trustworthy Embodied Systems
A new generation of increasingly autonomous and self-learning systems, which we call embodied systems, is about to emerge. When deploying these systems into our very societal fabric, we face various
CitySpec: An Intelligent Assistant System for Requirement Specification in Smart Cities
This work builds CitySpec, the first intelligent assistant system for requirement specification in smart cities, and builds a translation model and enhance it through requirement synthesis and develops a novel online learning framework with validation under uncertainty.
An Intelligent Assistant for Converting City Requirements to Formal Specification
This paper presents City Spec, an intelligent assistant system for requirement specification in smart cities that helps overcome the language differences brought by English requirements and formal specifications, but also offers solutions to those missing, inaccurate, or ambiguous information.
Toward Formal Methods for Smart Cities
This talk discusses several core challenges and the recent efforts as the first step toward developing novel formal methods to ensure safety and performance in smart cities.


Bayesian Neural Predictive Monitoring
Neural Predictive Monitoring (NPM) is a technique that complements NSC predictions with estimates of the predictive uncertainty, and uses Bayesian techniques, Bayesian Neural Networks and Gaussian Processes to learn respectively the predictor and the rejection rule.
Safe Control under Uncertainty with Probabilistic Signal Temporal Logic
This work proposes the new Probabilistic Signal Temporal Logic (PrSTL), an expressive language to define stochastic properties and enforce probabilistic guarantees on them, and presents an efficient algorithm to reason about safe controllers given the constraints derived from the PrSTL specification.
STLnet: Signal Temporal Logic Enforced Multivariate Recurrent Neural Networks
This paper develops a new temporal logic-based learning framework, STLnet, which guides the RNN learning process with auxiliary knowledge of model properties, and produces a more robust model for improved future predictions.
Safe Autonomy Under Perception Uncertainty Using Chance-Constrained Temporal Logic
This paper proposes a probabilistic extension of temporal logic, named Chance Constrained Temporal Logic (C2TL), that can be used to specify correctness requirements in presence of uncertainty, and presents a novel automated synthesis technique that compiles C2TL specification into mixed integer constraints.
Stochastic contracts for cyber-physical system design under probabilistic requirements
We develop an assume-guarantee contract framework for the design of cyber-physical systems, modeled as closed-loop control systems, under probabilistic requirements. We use a variant of signal
Specification Mining and Robust Design under Uncertainty
The quantitative semantics for StTL is introduced to reason about the robust satisfaction of an StTL specification by a given system, and an algorithm for parameter inference for Parameteric-StTL specifications is formulated, which allows specifications to be mined from output traces of the underlying system.
Specification-Based Monitoring of Cyber-Physical Systems: A Survey on Theory, Tools and Applications
This chapter summarise the state-of-the-art techniques for qualitative and quantitative monitoring of CPS behaviours, and presents an overview of some of the important applications and describes the tools supporting CPS monitoring and compare their main features.
Adaptive Runtime Verification
A key aspect of the ARV framework is a new algorithm for RVSE that performs the calculations in advance, dramatically reducing the runtime overhead of RVSE, at the cost of introducing some approximation error.
Runtime Verification with State Estimation
This work views event sequences as observation sequences of a Hidden Markov Model, uses an HMM model of the monitored program to "fill in" sampling-induced gaps in observation sequences, and extends the classic forward algorithm for HMM state estimation to compute the probability that the property is satisfied by an execution of the program.
Accelerated Learning of Predictive Runtime Monitors for Rare Failure
This paper proposes a method of grammar inference by which a DTMC is learned with far fewer samples than normal sample distribution, and exploits the concept of importance sampling to construct accurate predictive models with orders of magnitude fewer samples.