A Secure Multi-Party Computation Framework for Decentralized Row Allocation in Cooperative Strawberry Harvesting Systems

This research introduces a cryptographically secure, fully decentralized protocol for coordinating harvesting robots in agricultural environments. A previously studied, partially decentralized, auction-based allocation mechanism is transformed to a privacy preserving, fully decentralized algorithm based on Secure Multi-Party Computation (MPC) and Shamir’s Secret Sharing (SSS). The proposed framework tackles the privacy vulnerabilities in multi-robot agricultural systems where sensitive operational parameters are exposed during coordination. Our methodology extends beyond the existing tri-layered architecture by eliminating the centralized arbitration requirement through another layer of peer-to-peer auction protocol that maintains strict privacy. Primary contributions include: (1) a decentralized auction mechanism ensuring parameter privacy, (2) a secure consensus protocol for resolving assignment conflicts without central coordination, and (3) formal privacy guarantees maintaining confidentiality of robot operational data. The experimental validation through Monte Carlo simulations demonstrates that the proposed privacy-preserving protocol achieves an equivalent assignment efficiency as the previously studied approach while providing cryptographic guarantees of parameter confidentiality. This work establishes a foundation for secure, coordination in agricultural robots where protection of operational parameters is crucial for commercial deployment.