Lightweight robots are frequently engaged in real-time tracking tasks to provide human companionship services. For effective target tracking, the YOLO series is often employed as a lightweight object detection framework in robot systems. However, YOLO still demands substantial resources to train larger-scale models, striking a balance between accuracy and resource efficiency. Deploying YOLO directly on robots with limited computing resources can lead to significant delays in detection, compromising the effectiveness of tracking tasks. A deeper concern arises from the prevalent use of CPUs as the primary computing units in robots, rendering many existing model optimization techniques, which primarily target GPU computing, unsuitable for this context.

Active random testing is a powerful technique to find concurrency bugs through predicting the potential buggy inter-leaves. It helps improve the effectiveness of random testing such that the buggy scenarios are selected actively from trivial ones. However, applying active random testing to find null pointer dereference (NPD) still faces a strong challenge in that these NPDS are usually caused by the nontrivial data races, and therefore it is insufficient to adopt a general dynamic prediction approach to find them. In this paper, we propose a race-driven active random testing approach, RADIATE, to detect NPDs. The essential idea of RADIATE is to perform a race-driven prediction of the original trace for obtaining the potential NPD scenarios, and then use active random testing technique to actively control the thread schedules for exposing the real NPDs. We have implemented our RADIATE approach, and evaluated it over 7 benchmark programs. The evaluation results show that RADIATE can effectively find the indiscoverable NPDs.