Source record

Published by arXiv cs.MA (T1 source). The original is at https://arxiv.org/abs/2607.14553.

Pipeline notes

The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.

Proposes an Intention Abstraction Layer (IAL) middleware for autonomous industrial systems. An LLM grounded in an OWL ontology parses natural-language goals into structured, persistent intention objects, with a consistency monitor that detects agent goal conflicts at registration time before execution. A proof-of-concept with two agents demonstrates pre-execution conflict flagging.

Addresses a concrete failure mode in multi-agent industrial control where conflicts surface only after missed targets or shutdowns. The LLM-plus-formal-ontology pattern for intention translation is a specific architectural idea worth tracking for agent coordination work.

Cited by

No citations on record.