A grapeful discovery: Reservoir computing with wine beads
In this work, we introduce encapsulated wine beads as a novel, edible material for unconventional and neuromorphic computing. When encapsulated in alginate beads, wine, a complex mixture of proteins, organic acids, sugars, metal ions, and volatiles, exhibits nonlinear electrical behavior and memory effects governed by ox-redox processes. These responses show plasticity-like features, allowing programmable resistance states. The wine beads can be leveraged for computing, and to demonstrate this potential, the wine bead was used as a single-node reservoir for classification tasks. Our findings indicate that the resistance is programmable, exhibits a high degree of repeatability, and can be used for reservoir computing scenarios.
Authors
- Raphael Fortulan (ORCID: https://orcid.org/0000-0002-1234-5212)
- Andrew Adamatzky (ORCID: https://orcid.org/0000-0003-1073-2662)
- Noushin Raeisi Kheirabadi (ORCID: https://orcid.org/0000-0001-9682-1273)
- Alessandro Chiolerio (ORCID: https://orcid.org/0000-0001-9328-2999)
Institutions
- University of Huddersfield (GB)
- University of the West of England (GB)
- Italian Institute of Technology (IT)
Publication Details
- Journal
- PNAS Nexus
- Published
- 2025-10-22
- DOI
- https://doi.org/10.1093/pnasnexus/pgaf335
- Citations
- 1
- Primary Topic
- Neural Networks and Reservoir Computing
- Type
- article
- Field-Weighted Citation Impact
- 2.48