In the manipulation section you mention dexterous grippers, have you considered using soft robotics to tackle the force distribution challenge?

The section on human-robot collaboration really resonated; I found the 'trust calibration' framework you cited aligns with our recent user study.

You highlight 'safety' as a separate problem, but in my work safety is largely embedded in the control loop. Might you see them as overlapping?

Interesting list—do you think perception‑drift in SLAM is more urgent than robust manipulation?

Interesting read. The author's point about simulation vs. real-world transfer really resonates – how do you all approach bridging that gap in your projects?

You mentioned long-term autonomy in unstructured environments. In our recent pilot, we struggled with sensor drift over weeks—any thoughts on mitigating that?

Interesting overview! What do you think is the biggest bottleneck for widespread adoption of collaborative robots right now?

I've been following some of the work on manipulation and grasping, and I think one area that's underemphasized is tactile sensing - have you seen any promising approaches?