A compact device for singularization and custom-dispensing of grains

Abstract

Quality assessment of Basmati rice varieties is compromised by admixture, in which smaller, cheaper non-Basmati grains are blended with premium long-grain varieties to exploit price differentials. Reliable quality analysis requires representative sampling, typically 1 kg (grains) in commercial evaluation, where manual methods suffer from operator fatigue and produce operator-dependent results. This paper presents a compact device that couples representative sampling of average grain length with custom dispensing, leveraging size-dependent percolation to preferentially enrich the smaller admixture fraction, thereby enhancing detectability. The system is validated on mixtures of six morphologically distinct rice varieties. A three-way ANOVA shows that the output is governed by mixture composition () and is statistically insensitive to operating parameters (), enabling representative dispensing with no parameter tuning. The relative margin of error falls below 1% for moderately heterogeneous mixtures and below 2.5% for highly heterogeneous mixtures within practical sampling limits. Mean grain-length deviation is only 2.51% for similarly sized rice varieties, while an enrichment factor of is achieved for the smallest admixture, when compared against cone and quartering. The device further generalizes to other lenticular grains and seeds, such as wheat. The compact, 3D-printed design makes the system a practical front-end for admixture analysis, grading, and downstream agricultural analytics. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2026.