RECIPE: A Vehicle RECall Incident PrEdiction Method

The automakers, being unaware of “Chink in the armor” - the vehicle component defects, could issue the recalls for the defective components, only after major incidents. It can cause substantial damage to both the passengers and business of the Auto Manufacturers. Thus, we present RECIPE, a vehicle recall incident prediction method which can predict the recalls with sufficient lead time. We demonstrate a huge efficiency gain compared to prior methods and showcase the robustness of RECIPE with its zero-shot learning capability.