Read article: https://arobs.com/blog/robotic-arm-how-to-create-an-intelligent-robotic-system-that-types-autonomously/
Basic concept of the project is based on the use of machine vision techniques, neural networks, positioning techniques in 3D space through the WEB camera, and, of course, the creation of a GUI for the interaction between the intelligent system and the user.  

The robotic arm will automatically cover the handwritten text. The text is acquired through a WEB camera after which it is sent to the server for interpreting and extracting the text. After the text is processed, the robotic arm equipped with another WEB camera detects the letters in real-time, from the keyboard. By using the deep cam algorithms, it traces the trajectory of the arm's movement from the current point to the required letter on the keyboard. This process is cyclically repeated for each letter in the word.