In Russia, Robots Learn to Detect Their Own Malfunctions
Researchers at Sevastopol State University have developed an innovative method for diagnosing complex robotic systems, allowing machines to identify their own malfunctions and distinguish root causes even when multiple failures occur simultaneously.
A research team at Sevastopol State University has developed a theory of diagnostics and fault-tolerant control for complex robotic systems.
The key challenge addressed by the method is separating symptoms from true causes when several defects arise at the same time. The researchers created a unique algorithm based on the parallel operation of an entire “bank” of specialized observer-detectors.
Each digital “diagnostician” is tuned to detect a strictly defined class of malfunctions, including mechanical impacts, electric motor failures, and sensor faults.
A Precise Diagnosis
The data are combined through a system of equations to produce an accurate picture of all damage occurring within the system. According to the researchers, this approach helps avoid false alarms and directs efforts toward eliminating the true causes of malfunctions.
