Low-cost robot overcomes almost any obstacle

A team of researchers has designed a robotic system that allows a low-cost, small-legged robot to navigate almost any obstacle or terrain. The robot can go up and down stairs almost at your height or navigate rocky, slippery, uneven, steep and varied terrain. You can also cross gaps, climb rocks, and operate in the dark.

the Project The system was developed by researchers from Carnegie Mellon University’s School of Computer Science and the University of California, Berkeley.

Empower little robots with new abilities

Deepak Pathak is an assistant professor at the Robotics Institute.

“Empowering small robots to climb stairs and handle a variety of environments is crucial to developing robots that are useful in people’s homes as well as in search and rescue operations,” Pathak said. “This system creates a robust and adaptable robot that could perform many everyday tasks.”

The robot was tested on uneven stairs and slopes in public parks, which tested its ability to walk on steps and slippery surfaces. He was also tasked with climbing stairs which would be the equivalent of a human jumping over an obstacle. The robot achieves an impressive ability to quickly adapt and dominate the terrain using its vision and a small on-board computer.

The robot was trained with 4,000 clones in a simulator. These clones practiced walking and climbing complex terrain, and the simulator’s speed allowed the robot to achieve six years of experience in a single day.

The simulator stored the motor skills learned during training in a neural network, which the researchers then copied to the real robot. This innovative approach meant that there was no manual engineering of the robot’s movements.

Many of today’s robotic systems are based on cameras that create a map of the surrounding environment, which is then used to plan the robot’s movements before they are carried out. However, this process can be slow and error prone due to inaccuracies or misperceptions at the mapping stage. These inaccuracies can affect planning and movements.

While mapping and planning are useful for systems focused on high-level control, they are not always best for the dynamic requirements of low-level abilities, such as walking or running.

Efficient and fast maneuvers

The newly developed robotic system skips the mapping and planning phases and directs vision inputs directly to the robot control. Basically this means that the robot sees and moves accordingly. The innovative technique allows the robot to react to its complex terrain very quickly and effectively.

The movements of the robot are trained through machine learning, making the robot low cost. The tested robot was at least 25 times cheaper than alternatives on the market. According to the team, their algorithm could make low-cost robots much more accessible.

Ananye Agarwal is an SCS Ph.D. machine learning student

“This system uses vision and feedback from the body directly as input to send commands to the robot’s motors,” Agarwal said. “This technique allows the system to be very robust in the real world. If you slip on the stairs, you can recover. He can go into unfamiliar environments and adapt.”

The robotic system was largely inspired by nature. For a robot the size of less than a foot tall, it has learned to adopt the movements humans use to climb over tall obstacles to scale stairs or obstacles of its height. The system uses hip abduction to overcome obstacles that are difficult for even the most advanced robotic leg systems available.

The team also looked to four-legged animals for inspiration.

“Four-legged animals have a memory that allows their hind legs to follow their front legs. Our system works in a similar way,” Pathak said.

Built-in memory allows the hind legs to remember what the camera saw, helping it to maneuver around obstacles.

Ashish Kumar is a Ph.D. student at Berkeley.

“Since there’s no map, no planning, our system remembers the terrain and how you moved the foreleg and translates that to the hindleg, doing it quickly and flawlessly,” says Kumar.

The new research could play an important role in solving some of the major challenges surrounding legged robots. It could even help lead to its use in homes.

Leave a Reply

Your email address will not be published. Required fields are marked *