Monday, August 4, 2025
HomeBusiness"He walks like a real wolf": this robot dog learns animal trips...

“He walks like a real wolf”: this robot dog learns animal trips in just 9 hours thanks to AI

In short
  • 🐾 Researchers at the University of Leeds have developed a four -legged robot that imitates movements Animals thanks to artificial intelligence.
  • This robot, nicknamed “Clarence”, learned to change Depending on the land in just nine hours.
  • Technology uses a frameworkDeep strengthening learning To simulate natural animal movements.
  • Potential applications include planetary explorationagriculture and infrastructure inspection, thanks to its capacity to adapt to various environments.

Technological advances in the field of robotics are constantly surprising, and researchers at the University of Leeds have recently taken a new step. By developing an innovative artificial intelligence system, they managed to provide a four -legged robot with the capacity to adapt its walking style to imitate that of a real animal, even on unknown terrain. This technological feat may well revolutionize the use of robots in dangerous environments, where human security is at stake.

Animal inspirations

To allow the robot to sail on land that he had never encountered before, researchers were inspired by the natural movements of various animals, such as dogs, cats and horses. These animals are experts in the art of adapting to different landscapes, adjusting their movements to save energy and maintain their balance. By observing these behaviors, scientists have designed a training framework allowing robots to switch from trotting to a race, or leap, just as mammals do.

According to Professor Zhou, the main author of the study, the objective was to provide robots with a strategic intelligence similar to that of animals, allowing them to instinctively adapt to their environment. By integrating principles such as balance, coordination and energy efficiency, researchers hope to develop embodied AI systems capable of moving and interacting with as much fluidity and resilience as living beings.

This giant robotic army becomes reality: Amazon reaches the symbolic course of the million machines in its logistics centers

Real world loan

One of the most remarkable aspects of this technology is the speed with which the robot, nicknamed “Clarence”, has learned to change its approach according to the field. In just nine hours, he mastered this skill, even surpassing many young animals. Thanks to a learning framework by deep strengthening, The robot can change your approach without any modification being necessary in the system. Joseph Humphreys, postgraduate researcher at the Leeds mechanical engineering school, stresses that all training courses take place in simulation, and that policies formed on computer can be transferred to the robot with intact efficiency.

The robot’s ability to adapt to changing environments, thanks to strategies inspired by animals, makes it ready for use in the real world as soon as it comes out of simulation. This eliminates the need for additional adjustments, a significant advance compared to current AI agents, often limited to learning a single task in a given environment.

An army of humanoid police robots in Indonesia: this major technological advance promises to drastically reduce crime and drug trafficking

The dog robot imitates nature

Biomimetic, or imitation of nature, is an promising solution to overcome complex mobility challenges. By imitating animals for navigation, engineers see new potential applications in planetary exploration, agriculture and infrastructure inspection. This experience also sheds light on the way in which biological intelligence can improve robotic systems, avoiding endangering animals for tests.

With the learning by deep strengthening, the robot was tested in hundreds of environments, such as wooden shavings, rocks, invaded roots and floating wood. Its objective was to adapt to different types of land by choosing the most appropriate approach. His ability to recover after sudden stumbles was tested by hitting his legs with a sweetening brush. Thanks to these experiences, it has demonstrated its ability to navigate effectively in various environments.

500 % more YouTube views: this formidable strategy propels influencers at the top of the online business

Towards robotics inspired by animals

Although tested only on a robot in the size of a dog, the system could be applied to other four -legged robots sharing a similar body shape. Methods inspired by animals can be adapted, regardless of the size or weight of the robot. This research, published in the journal Nature Machine Intelligence, opens up new perspectives for the future of adaptive robotics, where machines could imitate animal movement strategies to accomplish complex and varied tasks.

The development of robots capable of instinctively adapting to their environment poses an intriguing question: how will these advances transform our perception and our use of machines in our daily life?

This article is based on verified sources and the assistance of editorial technologies.

Did you like it? 4.5/5 (24)

tatum.wells
tatum.wells
Tatum’s Austin music column ranks taco-truck breakfast burritos alongside indie-band demos.
Facebook
Twitter
Instagram
RELATED ARTICLES

LEAVE A REPLY

Please enter your comment!
Please enter your name here

- Advertisment -

Most Popular

Recent Comments