The automated guided vehicle (AGV) is widely used in industrial environments for goods transportation. However, issues such as mechanical wear, reduced battery life, navigation error accumulation, and decreased operational efficiency caused by frequent starts and stops need to be addressed. This paper proposes an improved Autonomous Emergency Braking (AEB) algorithm to tackle these problems. The algorithm employs a stepwise deceleration strategy, effectively reducing the frequency of sudden stops and enhancing the system’s operational smoothness. The AEB algorithm not only considers straight-line driving scenarios but also optimizes deceleration strategies for turning scenarios, adjusting the deceleration detection range according to the turning trajectory. Additionally, a velocity smoothing algorithm is designed to ensure that speed changes during deceleration are gradual, avoiding abrupt speed variations that could impact the system. The feasibility of the AEB algorithm is validated through testing on actual equipment, and its performance is compared to that of a conventional emergency stop strategy. Experimental results show that the AEB algorithm significantly reduces the number of sudden stops, improves the AGV’s operational smoothness and safety, and demonstrates excellent adaptability and robustness across different operational conditions.