Fault Tolerance Tool for Human and Machine Interaction & Application to Civilian Aircraft

EasyChair Preprint 5050

Abstract

Enhancing human-machine interaction is critical to aerospace applications. An essential requirement in safety critical systems is the clear need to guarantee trustworthiness of a system as well as V&V (Verification and Validation). However, the current state of the art concerning decision support systems lacks effective tools in this area. The Coherence Function Package, introduced in this research, is a tool towards providing assurance that the action needed has the approval of both the human and the machine in terms of safety. These algorithms shed light on the future of an Explainable Artificial Intelligence (XAI, [1]), that fosters a synergy between these two factors. This vital requirement that has been further underscored after the tragic events of the Boeing 737 Max 8 crashes [2]. Preliminary results show that the proposed approach is not only able to detect any errors in the system, it also assists in circumventing conflicts leading to incoherence and suggests a preferred solution in real-time.

Keyphrases: Flight critical systems, aircraft safety, equivalence, human-machine interaction, implication, recursive function, tautology

@booklet{EasyChair:5050,
  author    = {Javier Viaña and Kelly Cohen},
  title     = {Fault Tolerance Tool for Human and Machine Interaction & Application to Civilian Aircraft},
  howpublished = {EasyChair Preprint 5050},
  year      = {EasyChair, 2021}}