Abstract: The advancement of intelligent mining technologies has made autonomous emergency braking (AEB) systems a critical component of mining transportation vehicles, aiming to ensure safety during operations. However, open-pit mining areas pose unique challenges that significantly impact the effectiveness of AEB systems. These environments are often defined by extreme weather conditions—such as rain, snow, and fog—that reduce visibility, as well as harsh operational factors such as dusty air, wet and slippery roads, steep slopes, and varying vehicle loads. These factors typically lead to delayed braking, resulting in collisions or unnecessary braking in otherwise safe conditions, both of which can compromise operational efficiency and safety. Addressing these functional safety challenges is crucial for advancing the reliability and adaptability of AEB systems in such dynamic scenarios. This study investigates the intended functional safety of AEB systems for mining transportation vehicles. Specifically, it focuses on mitigating safety issues due to system deficiencies triggered by high-risk scenarios. A comprehensive approach is adopted to analyze the causal factors behind unsafe system behaviors, with the objective of improving the perception and decision-making components. To systematically identify unsafe scenarios, a system-theoretic process analysis (STPA) method is employed. This method provides a structured framework for determining the root causes of unsafe behaviors within the AEB system. The analysis underscores key limitations in the perception system under adverse conditions, including inaccurate detection of obstacles due to environmental factors, and inadequacies in the decision-making system’s adaptability to the dynamic and challenging conditions of open-pit mining operations. For the perception component, an advanced image enhancement method based on the dark channel prior is implemented to enhance visibility in low-light and high-dust conditions. This method significantly improves the clarity of visual input, particularly in scenarios with poor lighting or heavy particulate interference. Additionally, an extended Kalman filter (EKF) fusion algorithm is employed to integrate data from multiple sensors, such as cameras, LiDAR, and radar, increasing the accuracy and reliability of the perception system under adverse weather conditions. For decision-making, a collision time model that considers critical environmental and operational factors, including road slope, surface adhesion coefficient, slip ratio, and vehicle load, is developed. This model ensures enhanced responsiveness and adaptability for the decision-making system under real-world conditions, improving its ability to make accurate and timely braking decisions. To validate the proposed improvements, a joint simulation platform is established by integrating MATLAB/Simulink, TruckSim, and PreScan. This platform allows comprehensive testing under hazardous scenarios through the selection of test cases representing the most critical risk conditions. Simulation results show that systems with enhanced perception alone still struggle in non-intersection scenarios, where delayed braking leads to collisions. Similarly, systems relying solely on improved decision-making tend to be overly conservative, triggering premature braking owing to inaccurate perception data in low-visibility or low-adhesion conditions. Conversely, systems with integrated improvements to perception and decision-making components demonstrate significantly enhanced performance. These systems effectively prevent collisions while minimizing unnecessary early braking, achieving a balance between safety and operational efficiency. This study provides a holistic approach to addressing the limitations of AEB systems in mining transportation vehicles. By comprehensively improving the perception and decision-making systems, the proposed solutions enhance the safety and adaptability of AEB systems, enhancing their suitability for deployment in complex and dynamic environments, such as open-pit mining. The findings offer valuable insights for the future design, optimization, and implementation of intelligent safety systems in challenging industrial settings.