Researchers from the School of Computer Science at Carnegie Mellon University and the University of California
It is important to give small robots the opportunityclimb stairs and cope with any environment. This way they will become really useful; will be useful for work both in people’s homes and for search and rescue operations, scientists explain.
Engineers specially programmed andtested the robot to see how it navigates uneven stairs and hillsides in public parks and other challenging environments. It turned out that he quickly adapts, relying on artificial vision and a small on-board computer.
The researchers trained the robot with 4,000 clones of it in a simulator, where they practiced walking and climbing. As the authors of the development explain, the machine gained all six years of experience in just one day.
Most robotic systems usecameras to create a map of the environment and use it to plan movements before executing them. This is a slow process that often leads to failures. Mapping and planning are useful in systems that focus on high-level control, but are not always suitable for the dynamic demands of low-level skills such as walking or running over complex terrain.
The new system bypasses all these steps and guidesvision input directly to the robot control. In simple terms, what a robot sees determines how it moves. This ultimately allows him to quickly respond to the changing circumstances of rough terrain and effectively navigate through it.
The robot the engineers used was at least 25 times cheaper than available alternatives. The scientists' algorithm could make inexpensive robots more accessible and efficient.
Read more:
The main theory of the origin of man was refuted: where did we come from
Results of first cancer drug trial published
8 billion people now live on Earth: is overpopulation threatening the planet?