How to capture an object, not by coordinates but by the object itself when recording a Keyword test?

You might be using "Record Low Level Procedure" mode while recording your script.



Try Normal Recording.



or



In case your objects are moving or height, width or position changing frequently at that time you will prefer Descriptive Programming. (Descriptive programming u need to create manually its time consuming process but most powerful and beneficial in your case )  



Thanks

Ravik