Autonomous Engineering Operations: Integrating AI, IT, and Electrical Systems for U.S. Industrial Resilience

Abstract

The capability of autonomous engineering operations is a real and quantifiable capability in the timeframe of 2020-2025 that is becoming a reality in the U.S. industrial environment, integrating AI driven decision-making, the convergence of IT/OT platforms, and modern versions of the electrical control system in a unified way to operate without constant human intervention. The study is a mixed-methods evaluation of the combined effect of autonomous engineering operations, AI integration and IT-electrical convergence on operational availability and industrial resilience in U.S. industrial companies. The study employed a sequential explanatory design: a cross-sectional quantitative survey (N = 310) of industrial operations practitioners in the U.S. across seven sectors (discrete manufacturing, process manufacturing, power generation, oil & gas, water & wastewater, transportation, and defense industrial base) followed by a semi-structured interview (N = 22) with maximum-variation sampling for sector, role, and AI-adoption maturity. Five hypotheses were quantitatively tested using OLS structural-model estimation with Sobel-tested mediation. Measurement-model reliability was strong (α range 0.86–0.90; CR range 0.80–0.87; AVE range 0.45–0.57) and structural-model fit was good (χ²/df = 1.36; CFI = 0.960; TLI = 0.954; RMSEA = 0.035 [0.026, 0.043]; SRMR = 0.042). The results indicate that autonomous engineering operations has the largest direct path to the industrial-resilience outcome (β = 0.330, p < .001); AI-integration has a positive direct path to the operational availability outcome (β = 0.205, p < .001); IT-electrical convergence has an independent path to the resilience outcome (β = 0.208, p < .001); and the path from AI-integration to industrial-resilience is partially mediated by cybersecurity & resilience controls (Sobel z = 4.01, p < .001; mediation ratio ≈ 28%). There was no support for autonomous-operations moderation of the IT-electrical → availability path (β_int = −0.033, p = .530). 4 out of 5 hypotheses were supported. The interview corpus was qualitatively themed, resulting in five themes: T1 the autonomy-trust gap, T2 IT-OT convergence friction, T3 AI-enabled predictive value, T4 workforce and skill reconfiguration, and T5 cyber-physical resilience trade-offs. Meta-inference suggests that there is high convergence between quantitative paths and qualitative mechanisms, with the dominance of AEO on industrial resilience being explained by the interview participants mainly through T1 (the trust-building period needed before autonomy delivers value) and T3 (the amplification effect of AI on engineering-reliability practices). The quantitative-qualitative integration also extends the model: systematically, the participants identify workforce reconfiguration (T4) as a precondition that is not directly captured by the quantitative constructs. Implications for autonomous-operations investment sequencing and for federal industrial policy are drawn.