Opportunities for Explainable Artificial Intelligence in Aerospace Predictive Maintenance

This paper aims to look at the value and the necessity of XAI (Explainable Artificial Intelligence) when using DNNs (Deep Neural Networks) in PM (Predictive Maintenance). The context will be the field of Aerospace IVHM (Integrated Vehicle Health Management) when using DNNs. An XAI (Explainable Artificial Intelligence) system is necessary so that the result of an AI (Artificial Intelligence) solution is clearly explained and understood by a human expert. This would allow the IVHM system to use XAI based PM to improve effectiveness of predictive model. An IVHM system would be able to utilize the information to assess the health of the subsystems, and their effect on the aircraft. Even if the underlying mathematical principles are understood, they lack an understandable insight, hence have difficulty in generating the underlying explanatory structures (i.e. black box). This calls for a process, or system, that enables decisions to be explainable, transparent, and understandable. It is argued that research in XAI would generally help to accelerate the implementation of AI/ML (Machine Learning) in the aerospace domain, and specifically help to facilitate compliance, transparency, and trust. This paper explains the following areas: Challenges & benefits of AI based PM in aerospace Why XAI is required for DNNs in aerospace PM? Evolution of XAI models and industry adoption Framework for XAI using XPA (Explainability Parameters) Discussion about future research in adopting XAI & DNNs in improving IVHM.

[1]  Jimeng Sun,et al.  RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism , 2016, NIPS.

[2]  Luciano Floridi,et al.  Why a Right to Explanation of Automated Decision-Making Does Not Exist in the General Data Protection Regulation , 2017 .

[3]  Goran Nenadic,et al.  Wind Turbine operational state prediction: towards featureless, end-to-end predictive maintenance , 2019, 2019 IEEE International Conference on Big Data (Big Data).

[4]  Jaime S. Cardoso,et al.  Machine Learning Interpretability: A Survey on Methods and Metrics , 2019, Electronics.

[5]  Alessandro Rinaldo,et al.  Distribution-Free Predictive Inference for Regression , 2016, Journal of the American Statistical Association.

[6]  Carlos Guestrin,et al.  Anchors: High-Precision Model-Agnostic Explanations , 2018, AAAI.

[7]  Oluwasanmi Koyejo,et al.  Examples are not enough, learn to criticize! Criticism for Interpretability , 2016, NIPS.

[8]  Franco Turini,et al.  Local Rule-Based Explanations of Black Box Decision Systems , 2018, ArXiv.

[9]  Geoffrey E. Hinton,et al.  Distilling a Neural Network Into a Soft Decision Tree , 2017, CEx@AI*IA.

[10]  Artur S. d'Avila Garcez,et al.  Logic Tensor Networks for Semantic Image Interpretation , 2017, IJCAI.

[11]  Avanti Shrikumar,et al.  Learning Important Features Through Propagating Activation Differences , 2017, ICML.

[12]  Klaus-Robert Müller,et al.  Explainable artificial intelligence , 2017 .

[13]  Ankur Taly,et al.  Axiomatic Attribution for Deep Networks , 2017, ICML.

[14]  Arthur H. A. Melani,et al.  Equipment failure prediction based on neural network analysis incorporating maintainers inspection findings , 2017, 2017 Annual Reliability and Maintainability Symposium (RAMS).

[15]  Bernhard Haslhofer,et al.  Predicting Time-to-Failure of Plasma Etching Equipment using Machine Learning , 2019, 2019 IEEE International Conference on Prognostics and Health Management (ICPHM).

[16]  Martin Wattenberg,et al.  Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV) , 2017, ICML.

[17]  Carlos Guestrin,et al.  "Why Should I Trust You?": Explaining the Predictions of Any Classifier , 2016, ArXiv.

[18]  L. M. Lieberman,et al.  What If … , 1983, Journal of learning disabilities.

[19]  Bob L. Sturm,et al.  Local Interpretable Model-Agnostic Explanations for Music Content Analysis , 2017, ISMIR.

[20]  M. Elliot.  What is Explainable AI , 2018 .

[21]  Sercan Ömer Arik,et al.  Decision Input Candidate database O utput Matching losses Sparsity regularization Prototypes Thresholding C onfidence Prototype label match ⍺ , 2019 .

[22]  Minsuk Kahng,et al.  ActiVis: Visual Exploration of Industry-Scale Deep Neural Network Models , 2017, IEEE Transactions on Visualization and Computer Graphics.

[23]  nasa,et al.  Research and technology goals and objectives for Integrated Vehicle Health Management (IVHM) , 2019 .

[24]  Agustí Verde Parera,et al.  General data protection regulation , 2018 .

[25]  Ray Pugh,et al.  Operations and Maintenance Best Practices--A Guide to Achieving Operational Efficiency , 2002 .

[26]  Cynthia Rudin,et al.  Deep Learning for Case-based Reasoning through Prototypes: A Neural Network that Explains its Predictions , 2017, AAAI.

[27]  Amina Adadi,et al.  Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI) , 2018, IEEE Access.

[28]  Svetha Venkatesh,et al.  $\mathtt {Deepr}$: A Convolutional Net for Medical Records , 2016, IEEE Journal of Biomedical and Health Informatics.

[29]  Scott Lundberg,et al.  A Unified Approach to Interpreting Model Predictions , 2017, NIPS.