Independent Researcher

Akul Swami is an independent researcher in runtime assurance for safety-critical edge ML systems, with nine-plus years of embedded software engineering experience in regulated industries. Site: akulswami.github.io. ORCID: 0009-0003-9549-5543.

Articles 1 - 1
Edge AI

When Correct Outputs Aren't Enough: Testing Edge AI for Hidden Timing Failures on Shared Silicon - Blog

June 01, 2026

Edge AI systems on shared-silicon platforms can produce stable, accurate outputs while violating real-time deadlines under realistic deployment contention, a failure mode that accuracy-based validation cannot detect. This article presents a same-hardware controlled experiment on an NVIDIA Jetson Orin Nano Super demonstrating the gap, and proposes a joint verification condition combining an output-stability metric with a tail-latency bound that engineers can apply to their own systems.

Articles 1 - 1