Towards Self-Explaining Digital Systems: A Design Methodology for the Next Generation
- 1 July 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
As digital systems get ever more complex, their behaviour may at times appear unfathomable. Users will only be prepared to accept this if they are convinced that the system does indeed work correctly. Thus, we argue the need for self-explaining systems: systems that are able to explain their behaviour, and the reasons for it. In this paper, we propose first steps towards a design methodology for such systems, and argue that beyond user acceptance, self-explanation also has other applications such as self-verification and reconfiguration. We propose a conceptual framework for self-explaining systems, discuss how to achieve completeness, and consider implementation aspects.Keywords
This publication has 18 references indexed in Scilit:
- Methods for interpreting and understanding deep neural networksDigital Signal Processing, 2018
- Towards lightweight satisfiability solvers for self-verification2017
- Self-Awareness in Systems on Chip— A SurveyIEEE Design & Test, 2017
- KOCL: Power Self- Awareness for Arbitrary FPGA-SoC-Accelerated OpenCL ApplicationsIEEE Design & Test, 2017
- Towards a methodology for self-verification2017
- Property mining using dynamic dependency graphs2017
- Verifying the structure and behavior in UML/OCL models using satisfiability solversIET Cyber-Physical Systems: Theory & Applications, 2016
- Verification of proofs of unsatisfiability for CNF formulas2003
- A Rigorous View of Mode ConfusionLecture Notes in Computer Science, 2002
- Safe, Untrusted Agents Using Proof-Carrying Code1998