Reproduce “Connected Dependability Cage: Run-Time Function and Anomaly Monitoring for the Development and Operation of Safe Automated Vehicles” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Iqra Aslam, Nour Habib, Abhishek Buragohain, Meng Zhang, Andreas Rausch, Vaibhav Tiwari, Mohamed Benchat
The Connected Dependability Cage proposes two runtime checks for automated-vehicle perception: a Function Monitor compares AI models on safety-relevant objects, while an Anomaly Monitor flags inputs unlike the training data. In a qualitative test, the monitors flagged a poster-classification mismatch and a lying pedestrian; after retraining on that scene, it was no longer flagged as anomalous. The fallback switch was not validated.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.