Synergistic Multi-Robot Knowledge Aggregation for Energy-Sustainable Embodied Autonomy

• Kaushal Thaker

  Author

The rapid convergence of artificial intelligence and robotics has accelerated the development of Embodied AI (EAI) systems capable of autonomous perception, reasoning, and interaction within physical environments. However, the large-scale deployment of such systems introduces substantial energy demands arising from computation, communication, and physical task execution. This paper investigates the energy implications of different learning paradigms in EAI and proposes collective learning (CL) as an energy-efficient framework for scalable robotic intelligence. A formal analytical model is developed to characterize skill acquisition dynamics, knowledge reuse, and inter-agent learning synergy across robotic populations. The study compares isolated learning, incremental learning, transfer-integrated learning, and collective learning using a simulation framework that evaluates total learning episodes and associated energy consumption. Results demonstrate that collective learning significantly reduces redundant exploration and accelerates convergence by enabling distributed knowledge sharing among agents. Simulation outcomes indicate that CL achieves nearly 75% lower total energy consumption compared to isolated learning while maintaining faster skill acquisition across increasing agent populations. The findings highlight collective learning as a promising paradigm for achieving sustainable, scalable, and cooperative embodied autonomy in future intelligent robotic ecosystems.

• PDF

Copyright (c) 2026 International Journal of Intelligent Systems and Data Science

This work is licensed under a Creative Commons Attribution 4.0 International License.