Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

WaveVerif: Acoustic Side-Channel based Verification of Robotic Workflows

Zeynep Yasemin Erdogan, Shishir Nagaraja, Chuadhry Mujeeb Ahmed, Ryan Shah

TL;DR

This work tackles the problem of verifying robotic actions after command execution in potentially compromised environments by leveraging acoustic side-channel analysis. It proposes a passive, hardware-free verification framework that uses external audio recordings to fingerprint movement signatures, employing four ML classifiers (SVM, DNN, RNN, CNN) trained on MFCCs and related features. The approach is validated on a uArm-based setup across axis movements and common workflows, demonstrating 80%+ baseline accuracy and robustness to variations in speed, distance, and microphone placement, with higher performance for CNN/DNN. The findings suggest acoustic fingerprints can enable real-time, low-cost verification in safety-critical settings, complementing sensor-based methods and enabling enhanced trust and transparency in networked robotic systems.

Abstract

In this paper, we present a framework that uses acoustic side-channel analysis (ASCA) to monitor and verify whether a robot correctly executes its intended commands. We develop and evaluate a machine-learning-based workflow verification system that uses acoustic emissions generated by robotic movements. The system can determine whether real-time behavior is consistent with expected commands. The evaluation takes into account movement speed, direction, and microphone distance. The results show that individual robot movements can be validated with over 80% accuracy under baseline conditions using four different classifiers: Support Vector Machine (SVM), Deep Neural Network (DNN), Recurrent Neural Network (RNN), and Convolutional Neural Network (CNN). Additionally, workflows such as pick-and-place and packing could be identified with similarly high confidence. Our findings demonstrate that acoustic signals can support real-time, low-cost, passive verification in sensitive robotic environments without requiring hardware modifications.

WaveVerif: Acoustic Side-Channel based Verification of Robotic Workflows

TL;DR

This work tackles the problem of verifying robotic actions after command execution in potentially compromised environments by leveraging acoustic side-channel analysis. It proposes a passive, hardware-free verification framework that uses external audio recordings to fingerprint movement signatures, employing four ML classifiers (SVM, DNN, RNN, CNN) trained on MFCCs and related features. The approach is validated on a uArm-based setup across axis movements and common workflows, demonstrating 80%+ baseline accuracy and robustness to variations in speed, distance, and microphone placement, with higher performance for CNN/DNN. The findings suggest acoustic fingerprints can enable real-time, low-cost verification in safety-critical settings, complementing sensor-based methods and enabling enhanced trust and transparency in networked robotic systems.

Abstract

In this paper, we present a framework that uses acoustic side-channel analysis (ASCA) to monitor and verify whether a robot correctly executes its intended commands. We develop and evaluate a machine-learning-based workflow verification system that uses acoustic emissions generated by robotic movements. The system can determine whether real-time behavior is consistent with expected commands. The evaluation takes into account movement speed, direction, and microphone distance. The results show that individual robot movements can be validated with over 80% accuracy under baseline conditions using four different classifiers: Support Vector Machine (SVM), Deep Neural Network (DNN), Recurrent Neural Network (RNN), and Convolutional Neural Network (CNN). Additionally, workflows such as pick-and-place and packing could be identified with similarly high confidence. Our findings demonstrate that acoustic signals can support real-time, low-cost, passive verification in sensitive robotic environments without requiring hardware modifications.

Paper Structure

This paper contains 23 sections, 3 figures, 16 tables.

Table of Contents

  1. Introduction
  2. Motivation and Goals
  3. Networked Robots
  4. Acoustic Side-Channel Features for Verification
  5. Threat Model
  6. Hypotheses and Goals
  7. Related Work
  8. Methodology
  9. Experimental Setup
  10. Acoustic Verification Features
  11. Operational Parameters and Evaluation Conditions
  12. Movement Dataset
  13. Dataset Preprocessing.
  14. Machine Learning (ML) Methods
  15. Evaluation
  16. ...and 8 more sections

Figures (3)

  • Figure 1: Teleoperated Robot Architecture.
  • Figure 2: Robot Environment for Acoustic Side Channel. The experimental setup consists of a uArm robot placed on a desktop, performing predefined movements, while a nearby microphone records acoustic emissions for side-channel analysis.
  • Figure 3: Depiction of Common Warehousing Workflows.