# Failure-averse Active Learning for Physics-constrained Systems

arXiv: https://arxiv.org/abs/2110.14443 · HTML: https://mlsanity.com/paper/2110.14443/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 57 references that are in the map.
- First posted on arXiv 27 Oct 2021.

## Builds on (most cited first, 15 of 15)

- [Integrating Scientific Knowledge with Machine Learning for Engineering and Environmental Systems](https://mlsanity.com/paper/2003.04919/) (2020)
- [Bayesian Optimization with Unknown Constraints](https://mlsanity.com/paper/1403.5607/) (2014)
- [Constrained Bayesian Optimization with Noisy Experiments](https://mlsanity.com/paper/1706.07094/) (2017)
- [Scalable Constrained Bayesian Optimization](https://mlsanity.com/paper/2002.08526/) (2020)
- [Stagewise Safe Bayesian Optimization with Gaussian Processes](https://mlsanity.com/paper/1806.07555/) (2018)
- [Predictive Entropy Search for Bayesian Optimization with Unknown Constraints](https://mlsanity.com/paper/1502.05312/) (2015)
- [Safe Exploration for Interactive Machine Learning](https://mlsanity.com/paper/1910.13726/) (2019)
- [Active learning of deep surrogates for PDEs: Application to metasurface design](https://mlsanity.com/paper/2008.12649/) (2020)
- [Iterative Surrogate Model Optimization (ISMO): An active learning algorithm for PDE constrained optimization with deep neural networks](https://mlsanity.com/paper/2008.05730/) (2020)
- [Physics-Informed CoKriging: A Gaussian-Process-Regression-Based Multifidelity Method for Data-Model Convergence](https://mlsanity.com/paper/1811.09757/) (2018)
- [Active Learning for Gaussian Process Considering Uncertainties with Application to Shape Control of Composite Fuselage](https://mlsanity.com/paper/2004.10931/) (2020)
- [Partitioned Active Learning for Heterogeneous Systems](https://mlsanity.com/paper/2105.08547/) (2021)
- [APIK: Active Physics-Informed Kriging Model with Partial Differential Equations](https://mlsanity.com/paper/2012.11798/) (2020)
- [Neural Network Gaussian Process Considering Input Uncertainty for Composite Structures Assembly](https://mlsanity.com/paper/2011.10861/) (2020)
- [Physics-Information-Aided Kriging: Constructing Covariance Functions using Stochastic Simulation Models](https://mlsanity.com/paper/1809.03461/) (2018)

## Built on by (most cited first, 5 of 5)

- [A Survey on Deep Active Learning: Recent Advances and New Frontiers](https://mlsanity.com/paper/2405.00334/) (2024)
- [Future Aware Safe Active Learning of Time Varying Systems using Gaussian Processes](https://mlsanity.com/paper/2405.10581/) (2024)
- [ADs: Active Data-sharing for Data Quality Assurance in Advanced Manufacturing Systems](https://mlsanity.com/paper/2404.00572/) (2024)
- [Adaptive Active Learning for Online Reliability Prediction of Satellite Electronics](https://mlsanity.com/paper/2603.09058/) (2026)
- [Multi-output Extreme Spatial Model for Complex Aircraft Production Systems](https://mlsanity.com/paper/2604.22548/) (2026)

## Abstract (arXiv)

Active learning is a subfield of machine learning that is devised for design and modeling of systems with highly expensive sampling costs. Industrial and engineering systems are generally subject to physics constraints that may induce fatal failures when they are violated, while such constraints are frequently underestimated in active learning. In this paper, we develop a novel active learning method that avoids failures considering implicit physics constraints that govern the system. The proposed approach is driven by two tasks: the safe variance reduction explores the safe region to reduce the variance of the target model, and the safe region expansion aims to extend the explorable region exploiting the probabilistic model of constraints. The global acquisition function is devised to judiciously optimize acquisition functions of two tasks, and its theoretical properties are provided. The proposed method is applied to the composite fuselage assembly process with consideration of material failure using the Tsai-wu criterion, and it is able to achieve zero-failure without the knowledge of explicit failure regions.