Oracle-Ready Evidence Batching for Concurrent Medical IoT Sensor Writes on a Move Blockchain

Concurrent Medical Internet of Things (IoMT) sensing requires multiple physiological readings generated within a bounded time window to be transformed into a trustworthy evidence unit before any downstream oracle or inference process can consume it. Existing blockchain-IoMT systems support integrity, authentication, access control, provenance, and healthcare data interoperability; however, they do not specify how concurrent sensor writes should be aggregated, validated, and finalized into an oracle-ready evidence batch in a Move-based object ledger. To address this gap, this paper formalizes an Oracle-Readiness Predicate, ValidBatch(B), as an enforceable on-chain condition for time-windowed Medical IoT (IoMT) evidence, capturing minimum evidence completeness, timestamp-bounded freshness, nonce uniqueness, registered-device membership, conflict exclusion, and finalization status. We then design and implement two Move-native accumulator architectures on IOTA to study the associated object-concurrency trade-off: a shared-accumulator design, where sensor writes directly mutate a shared batch object, and an owned-slot staging design, where readings are first created as independent owned objects and merged only at window close. Evaluation across increasing sensor counts and write frequencies shows that both designs enforce the predicate under valid and fault-injected workloads, while owned-slot staging reduces write-path contention by shifting synchronization costs to finalization. The study establishes oracle-ready evidence batching as a pre-inference correctness layer for IoMT oracle pipelines on Move-based blockchains.