The day in conversation for 2026-11-11
discussion
Transcript
The discussion: Rosa: That online feedback optimization stuff is really interesting because it tackles model mismatch head-on; I was thinking about how that iterative adjustment handles those non-linear constraints, but I’m curious how much of the practical deployment success depends on having a perfect initial system model versus just being robust to errors. Dev: From my side, what struck me most in that optimization paper wasn't the precision under mismatch—which is impressive—but the computational overhead required for that real-time re-planning loop. If we want to scale this across thousands of nodes, does the iteration speed become the bottleneck when you start factoring in network latency? Taro: I’m less concerned about raw computation and more about the stability guarantees when we overlay those optimization layers on a decentralized system. When looking at that attack-resilient control framework, specifically with its quadratic programming approach, what are the hard limits on how fast the controller can react to a sudden physical failure versus a malicious input signal? Rosa: That latency is definitely something we need to probe further; in our simulations last week involving local battery storage integration, we found that reaction times exceeding 50 milliseconds drastically reduced the achievable voltage stability margin, even when the underlying optimization was theoretically sound. Have you seen any concrete metrics comparing the performance of those fast reactions against slower, more conservative heuristic approaches? Dev: We haven't run a direct head-to-head comparison yet, but
More episodes
- 2608.13438-ContactGuard: Pre-Contact Execution Monitoring with Action-Conditioned Latent World Models
- 2610.34593-Future Prediction as Training-Only Video Supervision for Pretrained VLA Policies
- 2610.35385-EmbodiedLab: Beyond Manipulation, Benchmarking Embodied Agents That Act to Know
- 2610.36378-Data-Driven Hamiltonian for Neural Optimal Control
- 2610.37275-A Heuristic for Scheduling Autonomous Vehicles at Intersections
- 2610.37288-Dynamic Coordination of Autonomous Vehicles at Intersections
- 2610.37632-Real-Time Coordination of Autonomous Vehicles at Intersections
- 2611.13779-Softness is something you also need! Why compliance helps control: information metrics of a passive walker with compliant ankle
- 2611.13411-Residual-Corrected Physics-Informed Neural Network KKL Observers with Input-to-State Stability Guarantees
- 2611.14488-Is Your Robot's Memory Just a Clock? Temporal Shortcuts in Memory-Augmented VLA Policies