Notes from the Lab
Technical deep-dives, integration guides, and field reports from building Physical AI systems at scale.
Integrating SRV with NVIDIA Cosmos: World-Scale Simulation
How we're leveraging NVIDIA Cosmos to simulate robot cells at planetary scale—validating perception stacks before physical deployment.
Automating the Mahindra XUV 3XO: A Case Study in Live Inspection
A deep-dive into deploying vision-guided robots for real-time quality inspection on the XUV 3XO production line.
3D Depth Sensors vs CMOS: Choosing the Right Eye for your Robot
Structured light, ToF, or stereo vision? A practical guide to selecting the optimal sensing modality for your application.
Sub-30ms Latency: The Physics of Real-time Robot Guidance
Breaking down the latency budget: from photon capture to trajectory dispatch. Every microsecond counts.
Building the Robots Learning Lab (RLL) in Chennai
Inside our new 12,000 sq.ft. facility dedicated to imitation learning, synthetic data generation, and SOP validation.
gRPC vs MQTT: Orchestrating the SRV Edge Kit Architecture
Why we chose gRPC for synchronous control loops and MQTT for telemetry. A tale of two protocols.
Path Correction for Universal Robots: A Plug-and-Play Approach
Real-time trajectory adjustment without modifying the robot controller. How we achieve sub-millimetre corrections.
Decoding SOP Monitoring for Aerospace Quality Assurance
Vision-based verification of assembly procedures in aerospace manufacturing. Zero defects, full traceability.
Imitation Learning for Complex Assembly in Electronic Manufacturing
Teaching robots delicate insertion tasks through demonstration. From human guidance to autonomous execution.
2026 Forecast: The Rise of Autonomous Physical AI Agents
Predictions for the next 18 months: onboard intelligence, edge-native perception, and the death of cloud-dependent robotics.