MARCOS and ACM: Comparing Approaches to Machine Consciousness
How can consciousness-inspired frameworks enhance robotic autonomy? This paper by Daniela López De Luise and colleagues introduces MARCOS (Movement of Autonomous Robotics Codelet System), a robotic consciousness model designed to support adaptive indoor navigation and decision-making.
Robotic Consciousness: Evaluation of a Proposal, authored by Daniela López De Luise, Nelson Biedma, Lucas Martín Rancez, Leonardo Isoba, and David Trejo Pizzo, focuses on MARCOS, a system implementing CoFram, a framework inspired by LIDA and global workspace theories, to improve autonomous robot performance.
Key Highlights
- Consciousness Framework (CoFram): Provides a flexible framework for developing consciousness-inspired robotic models, emphasizing concept learning and adaptation.
- Dual-Loop Architecture: Utilizes a Real-Time Controller (RTC) for immediate decisions and a Robot Task Adviser (RTA) for strategic planning and obstacle navigation.
- Dynamic Knowledge Base: MARCOS learns and updates its internal map, adapting to unknown obstacles and enhancing decision-making capabilities.
- Codelet Mechanism: Employs codelets—specialized agents handling specific tasks—to process concepts and propose strategies dynamically.
Connection to ACM
The Artificial Consciousness Module (ACM) aligns with this study through:
- Adaptive Learning: ACM can integrate MARCOS’ dual-loop architecture to enhance adaptability in dynamic environments.
- Conceptual Frameworks: CoFram-inspired methods can inform ACM’s strategy for balancing short- and long-term goals.
- Agent-Based Modeling: The use of codelets complements ACM’s modular design for simulating emergent behaviors.
For a detailed exploration of the MARCOS system and its applications, access the full paper here.
