The day in conversation for 2026-11-10
discussion
Transcript
The discussion: Rosa: The drive toward unified control architectures seems ambitious when you consider the latency differences inherent in composing diffusion policies from disparate modalities; how does LAG-Fusion actually manage consistency when one stream is running significantly slower than another? Dev: I think the core challenge with MATE isn't just fusing data, but ensuring that the reference-frame rebasing rule provides guidance without introducing unacceptable jitter when combining, say, visual input with tactile feedback. What specific metrics are being used to quantify that consistency threshold in terms of control error? Taro: If we look at EvoPlan's approach to planning within an open-world scenario, it relies heavily on those symbolic constraints derived from demonstrations; but if the perception pipeline is struggling with asynchronous streams, how does the evolutionary search adjust its fitness function when the underlying state estimations are temporally inconsistent? Rosa: That inconsistency points toward infrastructure concerns; specifically regarding the Optical Network Digital Twin paper—does that twin model account for dynamic changes in signal quality that might directly impact the fidelity of those distance signals feeding into your planning models? Dev: The digital twin focuses on architectural requirements, which suggests a focus on quality awareness, but I’m curious if the
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