There are many tools in the software world that allow you to safely upgrade running systems without stopping them. Systems such as search engines, social networks, and streaming or multiplayer gaming platforms are almost never stopped when they are upgraded.
However, once we encounter systems running in the physical world, the task becomes very difficult, almost unsolvable. Imagine a smart city in which cars drive themselves, smart traffic lights regulate traffic, and automatic monitoring systems control data flows. And it's all controlled by artificial intelligence technology. How do you update the software on these autonomous agents so that nothing stops or fails?
JASS-2023 is dedicated to the study of smart context-sensitive traffic control and continuous software evolution technologies for cyber-physical systems. As part of the program, we will use the Duckietown model city and its self-driving cars and smart traffic lights (all guided by AI) to help us solve such problem. Several international student project teams will work together in Agile mode for a week, supported by professors from Technical University of Munich
, Imperial College London
, and Neapolis University Pafos
. Students will use real hardware equipment when working on projects. Core technologies:
Linux, ROS, Docker/Kubernetes, Matter/Thread, Duckietown, Python