Facebook Robot Assistants: A New Revolution

Facebook a multi-national company known to provide the best social networking platform in the world; has started building Reinforcement learning-based robots that can be self-capable of doing many real-world things. Facebook is already using the concept of Natural Language, Convolutional Neural Network, LSTM models, and many more but with the incorporation of Deep Reinforcement learning the company may again stand out to be one such organization that has proven to build AI-based robots that can perform real-world cases.

The idea behind the working of the robots created by Facebook’s FAIR laboratory is exploring a particular environment, creating the map of the environment by undergoing numerous hit and trials, and once the road map is created, moving according to the map with much more accuracy than humans.

Also, the map created by the robot utilizes the minimalistic distance and thus reaching the destination accurately as well as in a short period. The creation of a map and recognizing the same by the robot are mainly done with the help of two new algorithms that have been built by FAIR.

There is also a provision of sound waves that can be recognized by the robot, so that the robot has multi recognizing features and thus reaching the destination without any hindrance like a table coming in between or say a chair placed in the path of the robot. The research started in 2019 in June, and since then FAIR is trying its level best to achieve success.

The creation of the two algorithms and inculcation of sound was done in January 2020. The company provided the robot with just a depth-sensing camera, compass data, and GPS, and with these things, in hand, the robot was able to locate the shortest distance possible to reach the destination just like a man would have done. Also, the inclusion of sound was made by creating environments where sound waves were passed and based on the compression and rarefaction principle. When these sound waves hit the obstacles, the robot will be able to guess the hurdle present in its path and overcome the same.

The combination of sound and sight may help researchers to break the milestones in Reinforcement learning and make a very human-friendly robot that is capable enough to do various tasks like cooking, bringing grocery items from the store, ironing clothes, etc.

The AI laboratory FAIR is also trying to create different simulation environment where these bots will be trained on how to differentiate things with one another that is if somebody wants to make a cake then what are the items he/she would need to bake the same and then after recognizing the items what next can be done with the items so that the final output comes out to be a delicious cake.

These advancements in Reinforcement learning would aid in uplifting the economy of any country and in creating smart assistants that can help in fulfilling the needs of people. Also, with such advancements, there would be a time when we will achieve the point of Super Artificial Intelligence and help this world to be a better and smart place to live in.

Some Milestones in Reinforcement Learning that FAIR has given

• Real Deal.
• Home Alone.
• Mirror World.
• Uncommon Sense.

Conclusion

With such developments taking place in the world the workload of humans will get reduced to a great extent and also the efficiency of any person will increase thus enabling him/her to perform better in every aspect of life.