Lifelong Computing
In the future hyper-connected digital world, computing systems will need to work radically different than today, connected to a tremendous amount of diverse data and facing a priori unknown conditions during operation.
However, current computing systems are in essence built to realize a set of predefined tasks for a given context in which they operate. A multitude of techniques have been developed that enable computing systems to adapt dynamically with changing conditions, such as computational reflection, context-awareness, and online learning.
Lifelong Computing research addresses foundational challenges in the design and operation of long running computing systems operating in ever-changing environments.
Ever-running computing systems can be found on land (cars, cities, industries), in seas (sea farms, ships, offshore platforms) or in space (satellite, space observatories) and all are subjects to foundational paradigm changes for maintenance free operations.
To achieve lifelong computing systems, there is a need to develop a novel method for self-aware learning systems. Self-aware learning systems are capable to dynamically adapt their own hybrid learning architecture based on the novelty discovered by manifold learning and clustering. Lifelong dynamic learning architecture.
The scientists Thomas Bäck and René Vidal are leading figures in the field of unsupervised learning and evolutionary learning. They apply their knowledge at NORCE to develop Lifelong Computing algorithms to support evolution in systems and operations within different industries in Norway such as Offshore Oil & Gas, Ocean & Maritime and finally environment & Climate Change.
Research team at NORCE:
Xin Yao
Danny Weyns
Thomas Bäck
René Vidal
Nabil Belbachir
NORCE Norwegian Research Centre strive to be a leader in deriving research and technology development towards sustainable societies, economies and environments.
Outcomes of lifelong learning research is a leap in the intelligence of computing systems through the principles of lifelong modeling and understanding the world, perpetual learning, and automatically synthesizing new models and evolving the learning system to solve new problems.
Thus, lifelong learning would be an enabler for developing future systems that are able to ingest changes in its context, adapt/re-caliber and continue operating within the new settings.
More about Lifelong Computing
Read this article published in arXiv.org: Danny Weyns, Thomas Bäck, René Vidal, Xin Yao, and Ahmed Nabil Belbachir. 2021. Lifelong Computing. In Proceedings of ACM Conference (Conference’17). ACM, New York, NY, USA, 9 pages.
