Reflective parts expose a gap between recorded demonstrations and live grasp decisions.
The cell needs faster autonomous picks, while the policy must remain dependable on edge compute.
Who you’d be doing this for
“I watched it hover over three good parts today, then had to finish every pick myself.”
Xuan Prasert · Robotics Engineer
Owns uptime and throughput for the customer’s pick-and-place cell.
What is at stake
Reflective-part pick success is 82%, and operators are stepping in after visible targets trigger a pause. You have to weigh a targeted policy update against cycle time and intervention risk.
Why it isn’t already fixed
Every obvious fix costs something else. That’s the part you’d have to decide.
- pick success vs cycle time
- data coverage vs rapid iteration
- edge latency vs model robustness
- operator continuity vs policy experimentation
Why Applied Intuition
Dana’s robot learning loop is judged by whether policies improve on real industrial robots, not only in offline evaluation.
Written with these in mind
Not your kind of problem? 65 more at Applied Intuition, or browse every organization.
This is the setup. The work is inside.
Running it puts you in the room: the full situation and its constraints, stakeholders who push back in their own words, and the decisions that are yours to make. What you produce becomes a Day One Plan — work you can show someone instead of describing.