Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SUNSET -- A Sensor-fUsioN based semantic SegmEnTation exemplar for ROS-based self-adaptation

Andreas Wiedholz, Rafael Paintner, Julian Gleißner, Alwin Hoffmann, Tobias Huber

TL;DR

SUNSET tackles the challenge of evaluating architecture-based self-adaptation in robotics under concurrent runtime uncertainties with unknown sources. It introduces a ROS2-based exemplar featuring a sensor-fusion semantic-segmentation pipeline driven by a trained ML model, with uncertainty injection, a baseline managing system, and detailed documentation to enable reproducible, fair comparisons. The approach emphasizes symptom-to-source separation, supports multiple simultaneous uncertainties, and enables evaluation of self-healing and self-optimisation adaptations in a realistic UAV perception workload. By providing extensible architecture, uncertainty injection, and robust evaluation metrics, SUNSET offers a practical benchmark for advancing autonomic, ROS-based robotic systems to operate reliably in dynamic environments.

Abstract

The fact that robots are getting deployed more often in dynamic environments, together with the increasing complexity of their software systems, raises the need for self-adaptive approaches. In these environments robotic software systems increasingly operate amid (1) uncertainties, where symptoms are easy to observe but root causes are ambiguous, or (2) multiple uncertainties appear concurrently. We present SUNSET, a ROS2-based exemplar that enables rigorous, repeatable evaluation of architecture-based self-adaptation in such conditions. It implements a sensor fusion semantic-segmentation pipeline driven by a trained Machine Learning (ML) model whose input preprocessing can be perturbed to induce realistic performance degradations. The exemplar exposes five observable symptoms, where each can be caused by different root causes and supports concurrent uncertainties spanning self-healing and self-optimisation. SUNSET includes the segmentation pipeline, a trained ML model, uncertainty-injection scripts, a baseline controller, and step-by-step integration and evaluation documentation to facilitate reproducible studies and fair comparison.

SUNSET -- A Sensor-fUsioN based semantic SegmEnTation exemplar for ROS-based self-adaptation

TL;DR

SUNSET tackles the challenge of evaluating architecture-based self-adaptation in robotics under concurrent runtime uncertainties with unknown sources. It introduces a ROS2-based exemplar featuring a sensor-fusion semantic-segmentation pipeline driven by a trained ML model, with uncertainty injection, a baseline managing system, and detailed documentation to enable reproducible, fair comparisons. The approach emphasizes symptom-to-source separation, supports multiple simultaneous uncertainties, and enables evaluation of self-healing and self-optimisation adaptations in a realistic UAV perception workload. By providing extensible architecture, uncertainty injection, and robust evaluation metrics, SUNSET offers a practical benchmark for advancing autonomic, ROS-based robotic systems to operate reliably in dynamic environments.

Abstract

The fact that robots are getting deployed more often in dynamic environments, together with the increasing complexity of their software systems, raises the need for self-adaptive approaches. In these environments robotic software systems increasingly operate amid (1) uncertainties, where symptoms are easy to observe but root causes are ambiguous, or (2) multiple uncertainties appear concurrently. We present SUNSET, a ROS2-based exemplar that enables rigorous, repeatable evaluation of architecture-based self-adaptation in such conditions. It implements a sensor fusion semantic-segmentation pipeline driven by a trained Machine Learning (ML) model whose input preprocessing can be perturbed to induce realistic performance degradations. The exemplar exposes five observable symptoms, where each can be caused by different root causes and supports concurrent uncertainties spanning self-healing and self-optimisation. SUNSET includes the segmentation pipeline, a trained ML model, uncertainty-injection scripts, a baseline controller, and step-by-step integration and evaluation documentation to facilitate reproducible studies and fair comparison.
Paper Structure (12 sections, 3 figures, 1 table)

This paper contains 12 sections, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related work
  3. SUNSET
  4. Use-Case Description
  5. System Architecture
  6. Uncertainties
  7. Possible adaptations
  8. Expanding SUNSET
  9. Evaluation
  10. Metrics
  11. Baseline managing system
  12. Conclusion & Future work

Figures (3)

  • Figure 1: Overview SUNSET. Managed system and additional components are part of SUNSET. The managing system can be exchanged without adaptation in any of the ros nodes.
  • Figure 2: The effect of different uncertainties on the entropy of our segmentation model. This refers to symptom S4.
  • Figure 3: Potential adaptations in SUNSET. Blue := Reparametrization, Green := Change of communication, Orange := Activation, Deactivation and Redeploy