Robotic process automation

This invention describes a system for teaching a computer program, or "robot," how to automatically perform tasks that a human usually does on a computer. It works by observing a human user perform the task, capturing images or video of their screen using a camera, and then using computer vision to understand each step. From this visual observation, it creates detailed instructions that the robot can follow to do the task automatically.

Why it matters: Filed before the widespread adoption of advanced deep learning for computer vision and generative AI. Modern AI could significantly improve the accuracy of identifying activities and generating robust automation scripts from screen recordings.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedSoftware logic, computer vision, automation engine

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10