13 August 2015

Ask an Edge Network to design its next generation

Have a look at this interesting paper. Researchers led by the University of Cambridge have developed a robot that can autonomously designs and build other robots, test which one does best, and automatically use the results to inform the design of the next generation.

The abstract reads: Artificial evolution of physical systems is a stochastic optimization method in which physical machines are iteratively adapted to a target function. The key for a meaningful design optimization is the capability to build variations of physical machines through the course of the evolutionary process. The optimization in turn no longer relies on complex physics models that are prone to the reality gap, a mismatch between simulated and real-world behavior...

This makes me thinking a paper which I wrote recently “The Network is the Robot”.

Can you imagine developing an edge network, i.e. a robot, that can also autonomously designs and build other (softwarized) networks architectures, test which one does and adapts best, and automatically use the results to select the next generation ? Obviously I'm not yet talking about core networks, but dynamic capillary networks, or IoT...

A matter of defining sort of “cognition loops” (crunching sensed real world network big data) to generate a set of new virtual architectures to help efficiently exploring the next generation design space… And even more to analyze the influence of the physical resources and ambient constraints onto the diversity that can be achieved ! But it could be also some application social networks.

As they say in the paper “the key for a meaningful design optimization is the capability to build variations of physical machines”: that’s true also for softwarized architectures empowered with a proper level of cognition.