The International Institute of Information Technology Hyderabad’s (IIIT-H) Robotics Research Center (RRC) and Ati Robotics have collaborated for three years to develop AI-based navigation for autonomous warehouse robots. This research focuses on non-LiDAR-based technology, using cameras, image sequences, and 3D foundational models for navigation. The approach aims to reduce the need for heavy, dense LiDAR maps, employing sparse localization based on visual landmarks. Rather than using dense LiDAR maps, IIIT-H researchers experimented with sparse localization, which estimates a robot’s position with limited data or landmarks. In indoor settings like warehouses, this involves decomposing scenes into landmarks such as “next to the door” or “opposite the elevator,” according to Professor Madhava Krishna.

IIIT Hyderabad hosted the Digital Citizen Summit – Regional Dialogues: Digital Public Infrastructures for Governance on 28 October 2026, focusing on the evolution of India’s digital public infrastructure and facilitating its advancement. Collectively organised by Digital Empowerment Foundation (DEF), the Human Sciences Research Centre at IIIT-H and Code for Good Tech (C4GT), the regional dialogue brought together stakeholders from government, civil society, technology, research and community representatives who deliberated on making India’s digital public infrastructure more inclusive, accountable and accessible. India’s Digital Public Infrastructures (DPIs), including Aadhaar, the Unified Payments Interface (UPI) and DigiLocker, have transformed how citizens access public services, financial systems and welfare benefits.

The International Institute of Information Technology Hyderabad’s (IIIT-H) Robotics Research Center (RRC) and Ati Robotics have collaborated for three years to develop AI-based navigation for autonomous warehouse robots. This research focuses on non-LiDAR-based technology, using cameras, image sequences, and 3D foundational models for navigation. The approach aims to reduce the need for heavy, dense LiDAR maps, employing sparse localization based on visual landmarks. Rather than using dense LiDAR maps, IIIT-H researchers experimented with sparse localization, which estimates a robot’s position with limited data or landmarks. In indoor settings like warehouses, this involves decomposing scenes into landmarks such as “next to the door” or “opposite the elevator,” according to Professor Madhava Krishna.