Science Robotics
22 January 2020Vol 5, Issue 38
22 January 2020Vol 5, Issue 38
Data-driven approaches enable efficient large-scale autonomous navigation of wheeled-legged robots in urban environments.
A path-planning algorithm enables autonomous multidrone aerial surveys of Adélie penguin colonies in Antarctica.
A drone incorporating a force-sensing cage with adhesive surfaces enables environmental DNA to be collected from tree branches.
Electronic skin printed with nanomaterial inks enables machine learning–driven autonomous robotic physicochemical sensing.
The use of drones and artificial intelligence may offer more reliable methods of counting populations and monitoring wildlife.
Grasp analysis, motion plans, and field tests are presented for a robot using extendable booms to navigate difficult terrain.
A reinforcement learning–based controller enables real-world humanoid locomotion.
A micro-quadcopter with a passive telescopic leg shows boosted agility and endurance via hybrid locomotion.
OP3 humanoid robots learned to play agile soccer using deep reinforcement learning.
The science-fiction movie The Creator uses six real-world robots from the 1950s and 1960s to show progress in AI.