Train robots to detect human errors before they occur

A new ability for robots that humans don't have.

Researchers at Oklahoma State University are developing a system that allows burglaries to read signals from the human brain in real time and no, it's not conscious thoughts, but something even faster, a type of signal called ERRP, an electrical pattern that your brain and mine generate the instant we perceive an error, this happens before you even move your hand or take any action, in other words, the brain detects the problem first and now burglaries can do it too.

The system operates through a brain-computer interface, the operator uses a helmet with electroencephalogram sensors that captures these signals in real time, as soon as an error is identified, the robot can react immediately, reduce speed, stop or even return control to the human, all this in a matter of milliseconds, according to the researchers, this solves one of the biggest limitations of current systems, today a robot only perceives that something has gone wrong when it is too late, when it collides with an object or executes an incorrect action.

With this new approach, error can be anticipated before it occurs, which makes a big difference in high-risk environments such as nuclear facility operations, deep-sea exploration, or any scenario where a failure can have serious consequences. Another challenge was making the system adaptable, since each human brain is unique, to solve this, the researchers created a model that learns general patterns and then automatically adjusts to each user without the need for long calibration processes.

Perhaps the most interesting thing is in health.

A prosthesis that detects that you are executing an incorrect movement and adjusts automatically or an exoskeleton that responds to your instinct before you even manage to react, deep down, this technology points towards something much greater, an increasingly direct integration between humans and machines, where it is not just commands that control the robots, but signals from the brain itself.

And when this integration leaves the laboratory, it can be applied exactly where the risk is highest.