Archive 29.11.2023

Page 7 of 74
1 5 6 7 8 9 74

Using large language models to code new tasks for robots

You've likely heard that "experience is the best teacher"—but what if learning in the real world is prohibitively expensive? This is the plight of roboticists training their machines on manipulation tasks. Real-world interaction data is costly, so their robots often learn from simulated versions of different activities.

An approach that allows robots to learn in changing environments from human feedback and exploration

To best assist humans in real-world settings, robots should be able to continuously acquire useful new skills in dynamic and rapidly changing environments. Currently, however, most robots can only tackle tasks that they have been previously trained on and can only acquire new capabilities after further training.

A sensing paw that could improve the ability of legged robots to move on different terrains

Legged robots that artificially replicate the body structure and movements of animals could efficiently complete missions in a wide range of environments, including various outdoor natural settings. To do so, however, these robots should be able to walk on different terrains, such as soil, sand, grass, and so on, without losing balance, getting stuck or falling over.
Page 7 of 74
1 5 6 7 8 9 74