From materials to mission-ready autonomy.
"Advancing sensor-driven edge intelligence through device-aware architectures and AI-driven co-design, bridging nanoscale device physics, instrumentation, and system-level autonomy to deliver energy-efficient, secure, and dependable sensing for real-world challenges where intrinsic material behaviors become scalable, mission-ready technologies."
Prof. Santhosh Sivasubramani
Centre for CPS, IIT Hyderabad.
Indian Nanoelectronics User Program, IIT Bombay.
Device and architecture exploration using nanomagnetic logic and spintronics.
AI training and inference solutions, hardware co-design focus.
Design and development for avionics applications.
Development of hybrid sensor and processing integration for defense applications.
Research on approximate computing techniques for AI applications.
Collaborative research on cyber-physical systems and IoT technologies.
AI Hub for Productive Research & Innovation in electronics.
Leveraging industry platforms and partnerships (ARM, NVIDIA, Intel, AMD, Google Cloud, AWS) to explore cutting-edge technologies:
Developing tool-augmented agents using NVIDIA NGC NIM models, LangChain orchestration, and RAG architectures. Focus on multi-agent coordination, autonomous research assistants, and production deployment patterns.
Hardware acceleration research for TinyML and edge inference on ARM Cortex-M and Cortex-A processors. Exploring model quantization, custom instruction sets, and RTOS integration for real-time AI.
Next-generation telecommunication research using NVIDIA AI Aerial and Omniverse Digital Twin. Investigating intelligent RAN, spectrum optimization, and software-defined radio architectures for 6G.
Deploying NGC DeepStream SDK for intelligent video analytics pipelines. Applications in surveillance, autonomous systems, and industrial monitoring with edge-cloud processing.
Building reproducible MLOps workflows across Google Cloud and AWS infrastructure. Containerized training, automated deployment, and monitoring for production AI systems.
End-to-end chip design leveraging ARM processor IP, FPGA prototyping, and ASIC design flows. Targeting defense, aerospace, and critical infrastructure applications requiring indigenous technology.
Research enabled through industry partnerships with ARM, NVIDIA, Intel, AMD, Google Cloud, and AWS, providing access to cutting-edge platforms, processor IP, and cloud infrastructure for innovation and student training.
