Study on the SOTIF of the intelligent heavy-duty truck localization module

Abstract

Safety of the Intended Functionality (SOTIF) refers to reducing unacceptable risks due to design inadequacies or performance limitations of intelligent systems. The localization module is a crucial component of the intelligent heavy-duty truck (HDT) automated driving (AD) system. Ensuring the SOTIF of this module is very important for HDT. Therefore, this paper conducts an in-depth SOTIF analysis of the localization module by identifying the performance limitations and triggering conditions of the Inertial Measurement Unit (IMU), Light Detection and Ranging (LiDAR), and Global Positioning System (GPS) within the module. Based on this analysis, a SOTIF model for the localization module is proposed. Using the KAIST dataset, abnormal signal injection methods are employed to simulate abnormal data signals caused by sensor performance limitations. The designed SOTIF model is compared with the federated Kalman multi-sensor fusion localization module model. The results demonstrate that the SOTIF model outperforms the federated Kalman multi-sensor fusion model in terms of performance and meets the SOTIF requirements. This is significant for ensuring the SOTIF of the localization modules in AD systems.