This workshop is cancelled
We are sad to say that the workshop will not take place as part of RSS this year. Many of the invited speakers are unfortunately unable to attend the workshop in person. Since that affects the quality of the workshop, we decided to postpone it to another conference to ensure that the workshop will be productive with a good turnout. We will soon make an announcement which conference we will target next.
Recent advances in technology have led to a rapid growth of interest in applying robotic systems in safety-critical domains such as transportation (autonomous driving), medical fields (surgical robots), manufacturing (assembly-line robots), and space exploration (autonomous rovers). However, this comes with several challenges: modern robotic systems not only present complex dynamics and have to deal with uncertainty, but also include machine learning components, such as neural network controllers. Because of these challenges, ensuring the robustness and safety of modern robotics systems is still an open question that requires an interdisciplinary approach between robotics, control, formal methods, and machine learning.
This workshop aims to bring together researchers interested in the broad area of safe and verifiable autonomy, which includes experts from the robotics, controls, machine learning, cyber-physical systems, and logic communities, as well as researchers working in the emerging area of safe and trustworthy AI. We attempt to highlight recent advances in these communities, discuss open problems and main challenges, and lay out new research directions.
- How to quantify data uncertainty for the purposes of controls?
- How to build robust and reliable machine learning models?
- How can robots safely interact with humans?
- Role of logic and formal methods in machine learning and robotics?
- What are the available tools and benchmarks?
- How to define/agree on a set of standard benchmark problems?