Sara Behdad

What they work on

1. Develop a hybrid digital twin that combines physics-based models with data-driven surrogate models through Bayesian multi-fidelity fusion, balancing computing cost and prediction accuracy for remanufacturing of used products.
2. Create a hierarchical reinforcement learning framework that links strategic planning with operational decisions in remanufacturing, and evaluate it on a robotic disassembly testbed for consumer electronic products in varying conditions.
3. Advance AI pattern recognition and fractal geometry analysis to extract shape, distribution and density patterns from images of biological attachment mechanisms such as gecko feet, burrs and bird feathers.
4. Build deep learning models for 3D reconstruction of biological attachment structures, then optimize bio-inspired probabilistic fasteners and test prototypes in simulation and real robotic assembly experiments.
5. Develop graph attention network algorithms that estimate missing lifecycle inventory (LCI) data for new product designs by comparing them with similar existing designs whose LCI data are known.
6. Model the uneven, random lifespans of smartphones, tablets and laptops with Markov process-inspired simulations, and use multi-criteria reinforcement learning to generate best trade-off designs that weigh environmental and technical goals.
7. Design low-cost active RFID tags that store each device's material types, quantities and recommended recycling processes, giving recyclers easy-to-read information about electronic waste.
8. Build decision-making models that use device-level tag data and reverse-logistics constraints to quantify reusability, guide consumers, repairers and recyclers, and support producer-responsibility cost policies.

More about the research

Why it matters

This work could help electronics recyclers, remanufacturers and manufacturers recover more valuable materials and send less e-waste to landfills and burning. Tools that fill in missing lifecycle data could let product designers check environmental impact early and cheaply. Digital twins and nature-inspired fasteners could make robotic disassembly and assembly more practical. Device-level recycling data could support policies that assign recycling costs to the companies that make electronics.

Known grants

From NIH and NSF research grants. Labs often have other funding we can't see.

Closest labs

The school

University of Florida: 65 labs on PI Pages, $1.3B a year on research, 795 research doctorates a year.

More about University of Florida

Reach out

A start for your email. Fill in the brackets.