# Manufacturers face coordination gaps in multi-agent AI systems

Digest AI · Enterprise & Industry · published 2026-09-21T15:06:57Z

Canonical: https://digestai.news/story/manufacturers-face-coordination-gaps-in-multi-agent-ai-systems

## Summary

Large-scale manufacturers are increasingly deploying AI agents for specific tasks like predictive maintenance and scheduling, but these systems often operate in isolation. While the percentage of organizations scaling AI agents in at least one function rose from 27% to 40% year over year, most focus on IT or software engineering rather than coordinated plant-wide operations. Research cited in the piece indicates that 79% of multi-agent LLM system failures stem from coordination and specification issues, highlighting a critical gap between deployment speed and infrastructure readiness.

The primary barriers to effective collaboration include fragmented data schemas in legacy MES and ERP systems, a lack of standardized communication protocols, and insufficient human oversight. The article notes that while protocols like Google’s Agent2Agent (A2A) and Anthropic’s Model Context Protocol (MCP) are emerging to standardize agent interaction, manufacturers must also address trust and transparency. Without shared context and clear escalation paths, agents cannot make decisions that account for the wider plant state, leading to suboptimal or unsafe outcomes.

To resolve this, the author argues that manufacturers should prioritize safety and security metrics over efficiency gains. Success requires establishing a shared operational vocabulary, similar to ISA-95 standards, that agents can process. By validating single agents before scaling to complex networks and ensuring human oversight is integrated into the architecture, manufacturers can avoid the pitfalls of agent sprawl and achieve reliable, collaborative automation.

## Key points

- 79% of multi-agent LLM system failures stem from coordination and specification issues, according to 2025 research.
- Organizations scaling AI agents in at least one function rose from 27% to 40% year over year.
- Google’s Agent2Agent and Anthropic’s MCP protocols aim to standardize how independent agents communicate and access tools.

## Why it matters

Manufacturers risk operational inefficiencies and safety hazards if AI agents operate in silos. Solving coordination and data sharing is essential for scaling agentic AI beyond isolated pilots to plant-wide automation.

## Sources

1. [Collaboration Must Sit At the Heart of Manufacturing’s Multi-Agentic AI Approach. Here’s How.](https://unite.ai/multi-agent-ai-coordination-manufacturing) (Unite.AI, 2026-09-21)

## Cite

Digest AI, "Manufacturers face coordination gaps in multi-agent AI systems", 21 September 2026, https://digestai.news/story/manufacturers-face-coordination-gaps-in-multi-agent-ai-systems

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/manufacturers-face-coordination-gaps-in-multi-agent-ai-systems.json