- Caro Roca, Martí
- Brando Guillaumes, Axel
- Abella Ferrer, Jaume
Dual (DMR) and Triple Modular Redundancy (TMR), often with some form of diversity, are used in safety-critical systems to realize those functionalities at the highest integrity level providing fault detection and/or tolerance capabilities. Redundant executions are intended to provide bit-level identical results and, upon any mismatch, an error is assumed and recovery actions taken as needed.
In this paper, we note that many emerging AI-based functionalities are intrinsically stochastic (e.g., camera-based object detection), and hence, their correctness must be judged semantically, with room for variations across correct outcomes (e.g., confidence must be above a given threshold, but how much it exceeds the threshold is irrelevant). Building on this observation, we propose strategies to create DMR and TMR implementations of AI-based functionalities that bring, not only fault tolerance against random hardware faults, but also against AI model inaccuracies. Those strategies, which can be realized with software-only means and ported to virtually any computing platform, build on input data modifications affecting the inference computations, but not the expected semantic output (e.g., introducing some controlled changes in the input data). Moreover, we provide our solution in the form of an open-source tool for image and video processing aimed at facilitating the reproducibility of our evaluation results, and enabling others to use it and conduct further research on input transformations., The research leading to these results has received funding from the European Union’s Horizon Europe Programme under the SAFEXPLAIN Project (www.safexplain.eu), grant agreement num. 101069595. This work is also part of the project PLEC2023-010240 funded by MICIU/AEI/10.13039/501100011033. Authors appreciate the support given to the Research Group SSAS (Code: 2021 SGR 00637) by the Research and University Department of the Generalitat de Catalunya. M. Caro has received funding under grant PRE2022-102440 funded by MICIU/AEI/10.13039/501100011033 and, by ESF+., Peer Reviewed, Postprint (author's final draft)