High-integrity Software for Robotics and Cyber-physical Systems
With the increasing presence in society of robotic an IoT-enabled devices, the reliability of their software is another critical aspect to be considered. The software for robots interacting with humans, autonomous vehicles, and robotized devices controlling critical resources, to name a few, are bound to face stringent development and certification requirements in the near future if they have to achieve mass adoption.
The topic of software reliability spans all the technological stack. On the hardware side of things, the current prevalence of multi-core CPUs opens the door to more efficient computing and more complex algorithms, but also difficulties the application of traditional real-time scheduling methods, critical for safe and robust control of systems. Meanwhile, the booming success of computation-intensive algorithms like machine-learning and deep neural network techniques also poses new challenges, as the resulting trained policies may be difficult to test for every scenario or be susceptible to attacks via specially crafted inputs.
There are multiple approaches for the achievement of reliable software: formal verification, design and testing methodologies, automated software generation, scheduling techniques, and more. This session aims at attracting researchers in these topics with a particular focus on its application to robotics and intelligent systems.
Theory and practice of high-integrity systems
Software architectures for reliable systems
Formal methods and algorithm proving
Design and implementation of real-time and embedded systems
Achieving and assuring safety in machine learning systems
Safe and secure cyber-physical systems
Languages for the development of reliable, safe and maintainable software
Modelling languages for critical systems
Robustness in distributed systems
Alejandro R. Mosteo, Centro Universitario de la Defensa
Danilo Tardioli, Centro Universitario de la Defensa, Spain