Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{196398,
        author = {Akash Singh and Vikash Singh and Rudrendra Bahadur Singh},
        title = {A Comprehensive Review on Decentralized Tri-Model AI Frameworks for Privacy-Preserving Autonomous Robotics},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {5336-5344},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=196398},
        abstract = {Autonomous service robotics is transitioning from cloud-dependent systems to decentralized, edge-centric architectures to address critical challenges in data privacy, network latency, and operational reliability. While Large Language Models (LLMs) such as Llama 3 and LLaVA offer sophisticated reasoning, their deployment on resource-constrained edge hardware often results in prohibitive inference delays and memory exhaustion. This paper reviews a robust alternative utilizing Small Language Models (SLMs), specifically Gemma 3, optimized via quantization techniques for real-time local reasoning. 
The proposed architecture integrates a fully local perception loop consisting of OpenAI Whisper for robust speech-to-text processing and Piper for low-latency vocal synthesis, ensuring 100% data sovereignty. Furthermore, we evaluate a "Lean Perception" mapping strategy that utilizes kinematic parameters—speed, time, and direction—to construct environmental representations with minimal computational overhead compared to traditional SLAM. 
By analysing 25 key research works in the domains of edge intelligence and autonomous navigation, this paper provides a comprehensive blueprint for private, high-performance, and responsive home assistant robots that operate independently of cloud infrastructure.},
        keywords = {},
        month = {April},
        }