Agentic AI in Engineering and Manufacturing: Industry Perspectives on Utility, Adoption, Challenges, and Opportunities

A qualitative study of 30+ industry interviews reveals that agentic AI adoption in engineering and manufacturing is progressing cautiously, with near-term value concentrated in structured, repetitive tasks and data synthesis. Adoption barriers stem primarily from fragmented data infrastructures, legacy system integration challenges, and organizational gaps rather than model capability limitations, requiring robust verification frameworks and human-in-the-loop governance before higher-order automation can scale.

This research provides empirical grounding for understanding how agentic AI actually integrates into complex industrial workflows, moving beyond vendor claims to capture practitioner realities. The study's core finding—that technical capability alone does not drive adoption—shifts focus from model performance benchmarks toward infrastructure and trust architecture. Engineers and manufacturers face a cascade of practical constraints: legacy CAD/CAM systems lack API accessibility, enterprise data exists in fragmented, machine-unfriendly formats, and regulatory requirements mandate auditability and human oversight that current agentic systems struggle to provide transparently.

The research maps a staged utility progression from low-risk assistance (drafting, documentation synthesis) toward orchestrated multi-step automation, contingent on infrastructure maturation. This staged model reflects industrial risk aversion—manufacturers cannot tolerate reliability failures in production-critical workflows. The organizational barriers identified—AI literacy gaps, governance misalignment, and cultural heterogeneity—suggest adoption timelines extend beyond typical enterprise software rollouts.

For the AI development sector, this underscores that competitive advantage lies not in larger foundation models but in domain-specific integration, verification tooling, and API ecosystem development. Vendors addressing data harmonization, legacy system bridging, and audit-trail generation face immediate market opportunities. The emphasis on spatial and physical reasoning breakthroughs indicates that manufacturing-specific AI capabilities remain underdeveloped relative to language or vision domains, creating differentiation opportunities for specialized builders.

• →Agentic AI value concentrates in orchestrating multi-step workflows across fragmented tools rather than replacing individual tasks
• →Data infrastructure and legacy system integration pose greater adoption barriers than current model capabilities
• →Verification, auditability, and human-in-the-loop governance are mandatory for manufacturing adoption due to regulatory and safety requirements
• →Organizational factors—AI literacy, cultural resistance, and governance misalignment—rival technical barriers in slowing deployment
• →Domain-specific integration and verification frameworks present higher market value than general-purpose model scaling