MOGOX Unveils MogoMind AI Model at WAIC 2025

From July 26 to 29, the annual global event in the field of artificial intelligence, the 2025 World Artificial Intelligence Conference (WAIC 2025), was held in Shanghai. During the conference, MOGOX showcased key technologies and products that demonstrate the application of AI large models in the transportation sector, including the MogoMind AI model, which deeply understands the physical world, and an AI network that interacts with the physical world in real time. MogoMind, as the first AI model capable of deep comprehension of the physical world, became one of the most attention-grabbing AI technology applications at the conference. Unlike large models in the digital world, MogoMind can be seen as a real-time search engine for the physical world, serving as a key to understanding reality and a super entrance to the real world. By accessing real-time dynamic data from the physical world, MogoMind achieves global perception, deep cognition, and real-time reasoning and decision-making capabilities, allowing it to extract meaning from data, learn rules from experience, and make flexible decisions in various scenarios. This provides crucial support for the construction of AI network infrastructure, achieving real-time digital twin effects and roadside data applications, becoming the 'AI digital base' for efficient urban and traffic operations. MogoMind addresses two core issues in current AI: the lack of real-time perception of the physical world and the absence of a global cognitive system. Through an integrated device that covers all areas, MogoMind can capture massive heterogeneous data such as vehicle trajectories, speed changes, traffic flow, and pedestrian dynamics around the clock, quickly integrating and processing this data through fusion algorithms to provide a data foundation for intelligent analysis and precise decision-making. With its real-time cognitive understanding of physical information, MogoMind can not only recognize road conditions, traffic signs, and obstacles but also transform complex traffic environment information into understandable and executable intelligent decision recommendations, offering solutions for traffic management departments and travelers. In terms of traffic flow prediction, MogoMind utilizes traffic flow prediction models and capacity evaluation algorithms to perform real-time dynamic calculations of road capacity, considering various factors such as traffic volume, vehicle types, road geometric characteristics, and traffic signal timing. Using reinforcement learning techniques, it uncovers patterns and trends in traffic data to predict future traffic flow changes. MogoMind also provides services such as real-time route planning, digital twin, and warning reminders, seamlessly integrating with traffic devices and systems from different manufacturers and types, including road sensors, onboard terminals, and traffic management systems, to achieve unified management and collaborative processing of multi-source data. For automakers, MogoMind offers various access solutions, facilitating the integration of platform data for function adaptation and application development. Government departments, traffic management agencies, and automakers can find suitable application scenarios within MogoMind to achieve resource sharing and complementary advantages, promoting the integrated development of AI and the transportation ecosystem. Based on these capabilities, MogoMind takes on three major roles: the 'decision-making hub' for urban traffic, the 'all-purpose assistant' for vehicle operation, and the 'invisible foundation' for autonomous driving. In the field of traffic management, MogoMind enables traffic managers to easily grasp the overall operation of the city's traffic system, whether it is macro-level traffic flow regulation, micro-level optimization of individual intersections, or emergency handling of road incidents, all based on scientific decisions made from the fusion analysis of real-time dynamic data, achieving overall collaborative optimization of urban traffic management. In the transportation sector, MogoMind provides deep understanding and planning decision-making services for real-time information in the physical world, offering capabilities such as super-visual traffic condition alerts, dynamic optimal route planning, and real-time perception of blind spot risks, ensuring driving safety and improving travel efficiency. In the autonomous driving domain, MogoMind enhances the safety and reliability of autonomous driving technology through multi-source data fusion and continuous learning from long-tail scenarios, feeding back into the training of autonomous driving models. Leveraging the capabilities of the MogoMind model, MOGOX has launched several L4 level mass-produced autonomous vehicles, including RoboBus, RoboSweeper, and RoboTaxi, which are deeply integrated with global perception, deep cognition, and real-time reasoning capabilities, promoting the application of autonomous driving technology across multiple scenarios such as public transportation, urban sanitation, and unmanned retail. Among these, the autonomous bus MOGOBUS, equipped with the end-to-end 'MogoAutoPilot+MogoMind' system, features real-time perception of traffic environments, road data analysis, and autonomous decision-making in emergencies, successfully operating in 10 provinces across the country with a safe driving mileage exceeding 2 million kilometers and serving over 200,000 passengers. Facing the AI era, all industries deserve to be reinvented with AI. The MogoMind large model showcased by MOGOX at this year's World Artificial Intelligence Conference is not just a technological breakthrough but also a forward-looking definition of the future form of AI and transportation integration, providing solid technical support for urban management, transportation, and the autonomous driving industry, facilitating the transition of urban transportation from 'point intelligence' to 'global intelligence.' As an important platform for China's technological innovation facing the world, the 2025 World Artificial Intelligence Conference focuses on a 'more effective' landing orientation with the theme 'Intelligent Era, Common Prosperity,' emphasizing core technologies such as large models, computing power chips, and data platforms, as well as industry applications in intelligent driving, smart cities, financial technology, and intelligent terminals like humanoid robots and autonomous driving systems, systematically outlining the intelligent industry map from foundational innovation to holistic empowerment. Over 1,200 AI leaders from more than 30 countries and regions came together to vividly portray the integration of AI into the fabric of cities and the lives of ordinary people.