Prof. John Doyle seminar: Neuro-inspired control and systems level synthesis
Place: Lecture hall KC:F, Kemicentrum, Naturvetarvägen 14, Faculty of Engineering LTH, Lund University, Lund.
Contact: anders [dot] rantzer [at] control [dot] lth [dot] se
Save event to your calendar
John C. Doyle, Professor of Control and Dynamical Systems, California Institute of Technology gives a guest lecture at the Department of Automatic Control.
Title: Neuro-inspired control and systems level synthesis
Where: Lecture hall KC:F, Kemicentrum, Naturvetarvägen 14, Faculty of Engineering LTH, Lund University, Lund.
Recent years have unfortunately highlighted intrinsic and systemic unsustainability and fragilities in our society and technologies. Detailed mechanisms underlying these in immune, medical ,computing, social, legal, energy, and transportation systems are incredibly diverse, but all are enabled by shared universal features of their architectures, whose designs are largely ad hoc historical artifacts. We need to more systematically design architectures that produce more robust and sustainable systems, including allowing higher layer learning and lower layer efficiencies to contribute. Using sensorimotor control, I’ll sketch the basic architectural concepts of laws, layers, levels, diversity, and sweet spots and the new math necessary to connect them, including System Level Synthesis (SLS). Crucial hardware layer constraints on sparsity, locality, and delay limit system layer functionality, robustness, and efficiency, but proper layering can mitigate this via diversity-enabled sweet spots (DeSS). In addition to our layered brains, illustrative examples include all our tech nets, layered immunity augmented by medicine and policy, systemic legal fragilities and the US 14th amendment, cascading failures in energy systems, climate change, language and its hijacking in social media, and wildfire ecosystems. New theory and technology is promising if it can be properly deployed. There are even some encouraging but relatively unexplored animal models for social architectures that give some broader hope.