Overview

An internal R&D concept simulation, built in NVIDIA Isaac Sim, of a Self-Driving Lab where AI designs the next experiment, a robot arm fabricates the materials, and measurement results feed back into the AI in a closed loop.

Project Background & Purpose

Using perovskite solar cell research as an example, we verified whether the precise, repetitive experiments researchers once ran by hand — mixing precursor solutions, spin/blade coating, annealing, and measurement — could be automated through an AI-robot closed loop. Because these repetitive experiments consume a great deal of skilled researchers' time, we first validated, through simulation, a lab automation model that lets a robot and AI handle the repetition so researchers can focus on higher-value decisions like experiment design and interpretation.

Components

• Robot-arm-based precision process automation
• Simulation of precursor-solution mixing / spin/blade coating / annealing
• Bayesian Optimization-based AI for recommending experiment conditions
• NVIDIA Isaac Sim virtual validation environment

Process Steps

1. 1AI designs the next experiment condition based on prior results
2. 2The robot arm carries out material fabrication: precursor mixing → spin/blade coating → annealing
3. 3Measurement results from the fabricated sample are fed back to the AI
4. 4Bayesian Optimization recommends the next experiment condition, repeating the closed loop