6thSense

6thSense builds touch-aware demonstration datasets for dexterous robotics. Its wearable and egocentric capture rigs record eight synchronized modalities per episode, including tactile and pressure signals, first-person video, depth, articulated hand pose, motion dynamics, secondary camera views, dense labels, and success or failure outcomes.

The company handles hardware, synchronization, calibration, and packaging so robot-learning teams receive aligned, model-ready episodes rather than raw sensor dumps. Representative task families include dishwashing, laundry, floor cleaning, and beverage preparation.