Danny Weyns

Professor, Linnaeus University and Katholieke Universiteit Leuven

The Vision of Lifelong Computing

Computing systems are becoming as vital as water and electricity for our society. Yet, engineering long running computing systems that achieve their goals in ever-changing environments pose significant challenges.

Today, dealing with bugs, new requirements, or new technologies requires system evolution. While several steps in the process of system evolution have been automated, it remains in essence a human-driven activity. Given the growing complexity of computing systems and the vast amount of highly complex data to process, human-driven evolution will eventually become unmanageable. To that end, we put forward a new paradigm for the design, operation, and evolution of computing systems that we coin lifelong computing. Lifelong computing systems integrate computing and learning elements. When the system detects the need for evolution (for instance the system detects an anomaly or a stakeholder adds a new requirement to the system), a lifelong computing system activates an evolutionary self-learning engine that runs online experiments to determine how the computing-learning system needs to evolve to deal with the changes.

In this process, the system can integrate new computing-learning elements as needed. Such computing-learning elements are provided by computing warehouses, where engineers can offer new elements together with data sheets and usage guides that allow lifelong computing systems to find the elements and integrate them autonomously to evolve as needed. In this talk, we motivate the need for lifelong computing and we outline a blueprint architecture for lifelong computing systems that we illustrate for a future fish farming system. To conclude, we highlight key research and engineering challenges that need to be tackled to realise the vision of lifelong computing.

Photo: Katrien Janssens, imec-DistriNet KU Leuven