货拉拉算法再公开：AI保障司机安全及判责公正

Core Insights - The article discusses the implementation of an AI safety monitoring and liability algorithm by Huolala, aimed at enhancing safety for freight drivers and improving liability experiences [1][9]. Group 1: AI Safety Monitoring - Huolala's AI safety monitoring system has been developed over four years and can detect hazardous goods transport, illegal passenger transport, fatigue driving, and overloading in real-time [1][4]. - The AI system can intervene by canceling orders or sending alerts to drivers when violations are detected [2][4]. - After the full application of AI, the daily risk orders related to hazardous goods transport and illegal passenger transport have decreased by 30% [4][7]. Group 2: Liability Assessment - The AI liability assessment algorithm collects various data post-order cancellation to determine responsibility, automatically ruling drivers as not at fault for canceled orders due to violations [1][9]. - The time taken for liability assessment has been reduced from 72 hours to 48 hours, and the appeal rate from drivers has decreased by nearly 30% [9][11]. - The algorithm ensures a fair and efficient process by combining machine analysis with human oversight for complex cases [11]. Group 3: Driver Safety Education - Huolala has provided 43 safety courses over the past year, with over 17 million participants, achieving a course satisfaction rate of 95% [8]. - The company has invested 1.75 billion yuan in occupational injury insurance for new employment forms, covering 8.9 billion orders and assisting over 1,200 drivers with claims [8]. Group 4: Future Developments - The company aims to continue optimizing its algorithms based on driver feedback, striving for a safer and fairer working environment for freight workers [11].