Institution profile

National Institute of Technology, Warangal

Academic institutionasia · in
Official website
Research library9linked papers
Opportunities0open roles
Selected work

Representative Papers

Automated Assembly Instruction Generation from CAD Models Using Grounded Large Language Models: A Human-in-the-Loop Framework

Oct 08, 2026

This study addresses the reliance on manual labor, fragmented workflows, and lack of collaborative oversight in converting CAD models into assembly documentation. To overcome these challenges, this work proposes an automated assembly instruction generation framework grounded in large language models. The framework employs a five-layer architecture that leverages STEP parsing and a ProductGraph intermediate representation to achieve structured information extraction and deterministic topological sorting. By integrating constrained contextual generation with visual augmentation, it realizes an end-to-end pipeline, while incorporating rule-based verification and human-in-the-loop mechanisms to ensure engineering accuracy. Experimental results validate the effectiveness of this system, which represents the first unified closed-loop approach to seamlessly integrate CAD interpretation, sequence planning, instruction writing, and human supervision.

0 citationsRead paper

AuthGuard-R: Safety-Compliant Mission Hijacking and Dual-Gate Defense for LLM-Controlled Robots

Sep 25, 2026

This study addresses the threat of "safety-compliant task hijacking" in LLM-based robots, wherein actions remain physically safe yet violate user authorization. To this end, it formalizes this attack model and proposes MissionPAIR, an adaptive attack framework, alongside AuthGuard-R, a deterministic authorization layer. By constructing a dual-gate architecture that decouples safety from authorization, the system cryptographically binds robot actions to signed task contexts, accompanied by a formal proof of authorization soundness. Across 240 real-world trials involving multiple models, including Claude Haiku and Qwen2.5, the proposed defense successfully intercepted all 109 unauthorized actions and eleven categories of protocol-level attacks with zero false positives.

0 citationsRead paper

Hierarchical Identity-Based Signature with Designated Aggregator from Lattices

Jun 12, 2026

This work proposes the first lattice-based hierarchical identity-based designated aggregate signature (HIBS-DA) scheme to address the high computational complexity and substantial resource overhead associated with multi-user data authentication in hierarchical organizations. The scheme enables users at different hierarchy levels to generate individual signatures that can be aggregated into a single compact signature, significantly reducing communication and verification costs while preserving security, correctness, and unforgeability. By integrating lattice-based cryptography with hierarchical identity-based signatures and designated aggregation, this approach is well-suited for large-scale hierarchical environments such as universities, corporations, and government agencies, offering enhanced authentication efficiency and scalability.

0 citationsRead paper
Recent publications

Latest Papers

Automated Assembly Instruction Generation from CAD Models Using Grounded Large Language Models: A Human-in-the-Loop Framework

Oct 08, 2026

This study addresses the reliance on manual labor, fragmented workflows, and lack of collaborative oversight in converting CAD models into assembly documentation. To overcome these challenges, this work proposes an automated assembly instruction generation framework grounded in large language models. The framework employs a five-layer architecture that leverages STEP parsing and a ProductGraph intermediate representation to achieve structured information extraction and deterministic topological sorting. By integrating constrained contextual generation with visual augmentation, it realizes an end-to-end pipeline, while incorporating rule-based verification and human-in-the-loop mechanisms to ensure engineering accuracy. Experimental results validate the effectiveness of this system, which represents the first unified closed-loop approach to seamlessly integrate CAD interpretation, sequence planning, instruction writing, and human supervision.

0 citationsRead paper

AuthGuard-R: Safety-Compliant Mission Hijacking and Dual-Gate Defense for LLM-Controlled Robots

Sep 25, 2026

This study addresses the threat of "safety-compliant task hijacking" in LLM-based robots, wherein actions remain physically safe yet violate user authorization. To this end, it formalizes this attack model and proposes MissionPAIR, an adaptive attack framework, alongside AuthGuard-R, a deterministic authorization layer. By constructing a dual-gate architecture that decouples safety from authorization, the system cryptographically binds robot actions to signed task contexts, accompanied by a formal proof of authorization soundness. Across 240 real-world trials involving multiple models, including Claude Haiku and Qwen2.5, the proposed defense successfully intercepted all 109 unauthorized actions and eleven categories of protocol-level attacks with zero false positives.

0 citationsRead paper

Hierarchical Identity-Based Signature with Designated Aggregator from Lattices

Jun 12, 2026

This work proposes the first lattice-based hierarchical identity-based designated aggregate signature (HIBS-DA) scheme to address the high computational complexity and substantial resource overhead associated with multi-user data authentication in hierarchical organizations. The scheme enables users at different hierarchy levels to generate individual signatures that can be aggregated into a single compact signature, significantly reducing communication and verification costs while preserving security, correctness, and unforgeability. By integrating lattice-based cryptography with hierarchical identity-based signatures and designated aggregation, this approach is well-suited for large-scale hierarchical environments such as universities, corporations, and government agencies, offering enhanced authentication efficiency and scalability.

0 citationsRead paper