Diagnostic testing represents a massive industry. In 2025, the global market for testing semiconductors for defects is projected to reach $39 billion. The medical laboratory testing market is even larger, valued at $125 billion.
Both semiconductor and medical testing deal with intricate systems containing many components, explains Rohan Ghuge, assistant professor of decision science at Texas McCombs School of Business. Whether the subject is computer chips or human bodies, testing involves managing complexity.
Research from Texas McCombs proposes a method to speed up testing by removing some costly and unnecessary steps. The study titled Nonadaptive Stochastic Score Classification and Explainable Half-Space Evaluation was published in Operations Research.
Typically, testing follows a sequence: instead of checking every component, clinicians prioritize certain tests first. Each phase excludes some potential issues and guides the next steps. However, this method can be time-consuming.
"First, you might check the vital signs."
— Rohan Ghuge, Texas McCombs
This stepwise testing creates delays and inefficiencies that the new approach aims to address.
The new testing framework holds promise to reduce time and cost in complex diagnostic systems for both semiconductors and medicine.
Author's summary: A novel testing method developed at Texas McCombs could make diagnosing complex systems like chips and medical conditions faster and more cost-effective by cutting redundant steps.