Finnish scientists have presented an inexpensive robot with which you can test the enhanced learning system. The device does not need a lot of data – it learns by navigating the environment around it.
Researchers have tried several times over the past decade to use Reinforced Learning (RL) approaches, such as training robots to navigate in any environment and perform basic tasks. But so far, they have not been able to create affordable devices capable of driving research facilities.
Researchers from Aalto University in Finland and Ote Robotics have unveiled RealAnt, an inexpensive four-legged robot that can test and implement RL algorithms. The new platform is a minimalist and affordable real-world version of the robot simulation environment used in reinforced learning research.
It was originally thought that algorithms would only work well if they were trained in simulators for thousands of hours. However, the researchers were able to train the four-legged robot to walk using minimal data. Therefore, the robot can be trained in a real environment without training in the simulator.
The researchers’ main goal was to create a simple and inexpensive robotic platform based on existing basic reinforced learning solutions. Such a platform will allow more researchers to build and test autonomous robots capable of performing basic tasks in the real world.
RealAnt, the four-legged robot they created, is versatile, minimalist, and inexpensive. He can independently learn to walk, move his legs in concert, and feel his position and orientation in a given environment. Using RL algorithms, a robot can be trained to perform simple yet valuable tasks. Moreover, for this, he focuses on the environment around him and does not need additional data.