Bengaluru Startup Is Building Wearable AI Cameras to Train the Next Generation of Humanoid Robots
India’s AI hardware ecosystem is quietly making waves on the global stage. Bengaluru-based startup Panoculon Labs is developing an innovative wearable camera system designed to help train humanoid robots by capturing human actions from a first-person perspective.
With lightweight hardware and advanced motion tracking technology, the company is creating tools that could play a key role in improving how robots learn to perform everyday tasks.
What Is Panoculon Labs?
Based in HSR Layout, Bengaluru, Panoculon Labs is an AI hardware startup founded by Sreeraj R. and Rishabh Sharma. The company specializes in wearable camera systems that record human movements from the wearer’s point of view, generating high-quality training data for humanoid robots.
As industries around the world invest heavily in robotics and artificial intelligence, accurate real-world data has become one of the most valuable resources for teaching robots how to interact with their surroundings.
A Three-Camera Wearable System
Panoculon Labs’ flagship product is a lightweight wearable kit featuring three synchronized cameras.
The setup includes:
One camera mounted on the user’s head.
Two cameras positioned near the wrists or palms.
This arrangement allows the system to continuously capture hand movements and finger positions, even when the hands move outside the field of view of the head-mounted camera. The result is a far more detailed representation of how humans naturally perform tasks.
Designed for Comfort During Long Sessions
One of the standout features of the wearable system is its remarkably low weight.
At approximately 35 grams, the device is significantly lighter than traditional recording setups that rely on action cameras or smartphones, which can weigh between 200 and 300 grams. The lightweight design helps reduce fatigue, allowing users to record data comfortably over extended periods.
For AI researchers, this means longer recording sessions and more consistent training data.
Why First-Person Data Matters
Humanoid robots learn by observing human behavior. Recording actions from a first-person perspective enables AI systems to better understand how people grasp objects, manipulate tools, and perform everyday activities.
The synchronized camera system captures precise hand and finger movements, providing richer datasets for robotics developers working on automation, warehouse operations, manufacturing, healthcare, and household assistance.
Advanced Tracking and Synchronization
Beyond its compact design, the wearable platform incorporates advanced synchronization and motion-tracking capabilities to ensure all three cameras record movements with high accuracy.
This synchronized data helps create detailed training datasets that are essential for improving robotic perception, object manipulation, and human-like movement.
Reaching Global Customers
The startup has also attracted attention beyond India, with its wearable AI camera systems reportedly being supplied to robotics companies working on humanoid technologies. This reflects the growing international demand for specialized hardware that supports AI model training.
Why This Innovation Matters
As humanoid robots continue to evolve, the quality of the data used to train them will become increasingly important. Panoculon Labs is addressing this challenge with a practical hardware solution that combines lightweight design, precise tracking, and first-person data capture.
The company’s work highlights how Indian startups are contributing not only to software innovation but also to cutting-edge AI hardware. If the global robotics market continues to expand, technologies like these could become an essential part of how intelligent robots learn to understand and interact with the real world.
