Meta has launched the PARTNR program, dedicated to exploring human-robot interaction, particularly emphasizing collaboration between humans and robots in domestic environments. The program targets household tasks like cleaning, cooking, and package delivery, and has unveiled a benchmark dataset consisting of 100,000 tasks designed to train embodied AI models.

Simulation technology is gaining significant importance in robot deployment. Meta's PARTNR model has successfully undergone testing on Boston Dynamics' Spot robot, alongside the development of a mixed reality interface that visualizes the robot's decision-making process. Furthermore, human-robot collaboration technology holds promise in elderly care, yet technological advancements are essential for widespread adoption.

Humanoid robots, a pivotal direction in human-robot collaboration, currently face challenges of high cost and the need for enhanced reliability. Consequently, manufacturers frequently prioritize meeting business demands over other considerations.