Designing a sensor's physical shape at the same time as its software increases accuracy by over 100 times.
April 29, 2026
Original Paper
Adaptive Sensing beyond Non-Adaptive Information Limits: End-to-End Co-Design of Geometry, Policy, and Inference
arXiv · 2604.25193
The Takeaway
Hardware and software are usually designed in isolation, which limits the potential of the final system. This project used joint dynamic programming to co-design the geometry of a sensor and its measurement policy. The result was a 123x improvement in sensing accuracy compared to standard baseline methods. By treating the physical structure as a variable in the optimization, the researchers found configurations humans would never have considered. This holistic approach could revolutionize how we build everything from medical imaging devices to autonomous vehicle sensors.
From the abstract
Inverse design has made vast physical parameter spaces a substrate for emergent behavior. In sensing, the stakes of this principle are sharpest at the analog-to-digital boundary, where any information the hardware fails to capture is information no downstream algorithm can recover; hardware optimization alone is therefore not enough, and the geometry must be co-designed with a rule for what to measure next. We formulate this co-design as \emph{joint dynamic programming} (joint-DP): a single opti