Software Quality Journal - CfP: Quality of Learning-enabled Autonomous Systems
Learning-enabled Autonomous Systems (AS) are nowadays largely adopted in various application domains (Autonomous Driving, Internet of Things, Web Services, Mobile Applications).
Emerging standards (e.g., ISO/IEC DIS 5338, ISO/IEC DIS 25059) confirm the need for novel contributions concerning the AS lifecycle and related quality assessment and improvement techniques.
This special issue solicits novel contributions in fundamental and applied research about the quality assessment and improvement of such systems, dealing with the most challenging issues they currently face:
- Integrate software engineering practices into the Autonomous Systems lifecycle according to paradigms such as MLOps.
- Measure Autonomous Systems quality (including reliability, safety, and security).
- Leverage monitored (unlabeled) data for quality assessment and improvement.
- Apply continuous quality assessment and/or improvement actions before and after the release in operation.
- Understand which actions can be automated and how.
Topics of interest concern quality assessment and improvement methods throughout the Learning-enabled Autonomous Systems (AS) lifecycle, including but not limited to:
- Quality attributes (e.g., performance, reliability, safety, security) impacting AS
- The role of Data Scientists, Machine Learning Engineers, and/or Domain Experts in quality assessment and improvement for AS
- Quality threats (e.g., the Oracle problem, incorrect design, unbalanced data)
- Quality of data (e.g., training, testing, operational data)
- Quality of the training process
- Quality of the AS models
- Quality assessment through testing
- Quality assessment through monitoring
- Quality prediction/forecasting
- Quality improvement of AS
- Automation of lifecycle steps (e.g., data processing, learning, testing, deployment, monitoring, and so on)
- Quality of AS for Image classification
- Quality of AS for Autonomous Driving
- Quality of AS for AIoT
- Operational context elicitation
- Solutions for green AS
- Social and ethical aspects