FLIP: Flowability-Informed Powder Weighing
Nikola Radulov, Alex Wright, Thomas Little, Andrew Ian Cooper, Gabriella Pizzuto
AI summary
Problem
Robotic automation of powder weighing struggles with sim-to-real gaps and poor generalization across diverse, cohesive powders due to complex, variable material dynamics.
Approach
FLIP uses automated angle of repose measurements to calibrate granular simulations via Bayesian optimization, then trains a reinforcement learning policy with a flowability-ordered curriculum learning strategy.
Key results
- Achieved 2.12 ± 1.53 mg dispensing error on real powders
- Outperformed domain randomization methods by reducing weighing error
- Successfully generalized to unseen, highly cohesive powders and new target masses
- Validated automated flowability measurement system against manual methods
Why it matters
Enables reliable, adaptive robotic powder handling for autonomous laboratory automation and materials discovery.
Abstract
Autonomous manipulation of powders remains a significant challenge for robotic automation in scientific laboratories. The inherent variability and complex physical interactions of powders in flow, coupled with variability in laboratory conditions necessitates adaptive automation. This work introduces FLIP, a flowability-informed powder weighing framework designed to enhance robotic policy learning for granular material handling. Our key contribution lies in using material flowability, quantified by the angle of repose, to optimise physics-based simulations through Bayesian inference. This yields material-specific simulation environments capable of generating accurate training data, which reflects diverse powder behaviours, for training ‘robot chemists’. Building on this, FLIP integrates quantified flowability into a curriculum learning strategy, fostering efficient acquisition of robust robotic policies by gradually introducing more challenging, less flow- able powders. We validate the efficacy of our method on a robotic powder weighing task under real-world laboratory con- ditions. Experimental results show that FLIP with a curriculum strategy achieves a low dispensing error of 2.12 ± 1.53 mg, outperforming methods that do not leverage flowability data, such as domain randomisation (6.11 ± 3.92 mg). These results demonstrate FLIP’s improved ability to generalise to previously unseen, more cohesive powders and to new target masses.