The Good, the Bot, and the User: Hybrid Automation to Harmonize Performance and Human Factors at the Intersection of Adaptive and Adaptable Systems

As automation increasingly supports complex decision-making, balancing system performance with human factors is a central challenge. Flexible automation is promising, yet the optimal configuration—adaptive, adaptable, or hybrid—remains debated and alternatives are seldom compared. This study compared adaptive (system changes automatically), adaptable (user configures the system), and two hybrid forms: hybrid-override (system-driven with user override) and hybrid-recommendation (system-suggested, user-approved). Participants completed a dynamic traffic control task under each condition. Subjective and objective measures were collected, including perceived autonomy, competence, role perception, preference, satisfaction, workload, situational awareness, and operator and system performance. Results confirmed a tradeoff: adaptive automation yielded the highest performance, while adaptable automation enhanced perceived autonomy, competence, and satisfaction. Hybrid approaches offered a middle ground but differed: hybrid-override leaned toward efficiency, whereas hybrid-recommendation better preserved user experience. Notably, hybrid automation is not a monolith; design variations shaped outcomes differently, motivating an integrative framework for hybrid automation design.