COMPARATIVE EFFICACY OF RULE-BASED RPA VS. COGNITIVE AI AGENTS IN ENHANCING OPERATIONAL EFFICIENCY

Abstract

This study investigates the comparative efficacy of rule-based Robotic Process Automation (RPA) and cognitive AI agents in enhancing operational efficiency, examining how process complexity moderates the effectiveness of each technology. A quantitative experimental design was employed using a controlled business process simulation with 199 participants randomly assigned to conditions varying by automation type and process complexity. Operational efficiency was measured through process cycle time, error rate, and cost per transaction. Two-way ANOVA results revealed a significant interaction between automation type and process complexity, F(1, 195) = 25.420, p < 0.001, supporting the central hypothesis. Rule-based RPA demonstrated superior efficiency for low-complexity, structured tasks (M = 59.17, SD = 248.06), while cognitive AI agents outperformed for high-complexity, knowledge-based processes (M = 43.21, SD = 9.83). The main effect of automation type was non-significant, indicating that neither technology universally excels across all contexts. The model explained 16.5% of the variance in operational efficiency. These findings contribute empirical evidence to the contingency perspective on automation selection, demonstrating that organizations must strategically align technology capabilities with process characteristics rather than adopting uniform automation strategies. The study addresses a critical gap in the literature by providing controlled comparative analysis of these automation paradigms. Practical implications suggest that firms should assess process complexity as a key determinant when making automation investments, deploying rule-based RPA for structured, high-volume tasks and cognitive AI agents for complex, knowledge-intensive processes. Future research should explore hybrid automation models and examine additional contextual factors across diverse industry settings.