Cognitively automated assembly processes: a simulation based evaluation of performance.

Area of Science:

• Robotics
• Artificial Intelligence
• Cognitive Science

Background:

• Robotic assembly cells require sophisticated control systems to handle complex tasks.
• Simulating human-like information processing in robots is key for adaptable automation.
• The SOAR cognitive architecture offers a framework for rule-based cognitive control.

Purpose of the Study:

• To evaluate the effectiveness of a cognitive control unit (CCU) based on SOAR for robot assembly.
• To investigate the performance of cognitive automation with deterministic parts supply, specifically addressing phase-shifts and part diversity.
• To determine if cognitive automation can efficiently manage planning problems with maximal part differences.

Main Methods:

• Development of a cognitive control unit (CCU) utilizing the SOAR cognitive architecture.
• Extensive simulation studies to test the CCU's planning and reaction capabilities.
• Analysis of processing times under different parts supply scenarios (random vs. deterministic) and part configurations.

Main Results:

• Cognitive automation with SOAR is highly effective for random parts supply, reducing logistical planning effort.
• Deterministic parts supply, particularly with numerous identical parts, led to increased processing times.
• The study demonstrates that cognitive automation is also suitable for deterministic parts supply with maximal part diversity, mitigating previous performance issues.

Conclusions:

• Cognitive automation, powered by SOAR, provides a robust framework for autonomous robotic assembly planning and execution.
• The CCU's adaptability extends to deterministic parts supply scenarios when dealing with diverse components.
• This research validates the applicability of cognitive automation for a wider range of industrial assembly challenges.