Human-Centered Interaction Technology in the Smart Cockpit Era
August 23 (Saturday) 3:45 PM-5:30 PM
Location: Grand Ballroom B, 3rd Floor

Introduction:
        Guo Gang is a professor and doctoral supervisor at the School of Mechanical and Vehicle Engineering, Chongqing University. He is also the chairman of the Intelligent Cockpit Branch of the China Society of Automotive Engineers (SAE-China) and a member of the board and expert committee of the China User Experience Alliance. His research focuses on the research and development of cutting-edge technologies and their application in the field, including intelligent connected vehicles, brain-cognitive mechanisms of human-machine interaction in intelligent cockpits, human factors engineering, user experience, ADAS/intelligent driving trust and takeover assessment, user experience testing and evaluation of intelligent cockpits, and key technologies and toolchain software for innovative design of scenario-driven intelligent cockpits. In recent years, as project leader or subject leader, he has undertaken three projects under the 13th Five-Year Plan National Key R&D Program, one automotive intelligent manufacturing special project of the Ministry of Industry and Information Technology, and two projects under the Chongqing Municipal Major Science and Technology Plan. He has also undertaken six commissioned projects from leading enterprises and received two second-class National Teaching Achievement Awards, one first-class provincial and ministerial-level science and technology award, and one second-class provincial and ministerial-level science and technology award. He has published over 40 high-quality academic papers in top SCI/EI journals, including one Nature journal (as corresponding author). He served as a consulting expert in the drafting of the "White Paper on Classification and Comprehensive Evaluation of Automotive Intelligent Cockpits" and spearheaded the development of two group standards, including the "Method for Evaluating the Intelligence Level of Automotive Intelligent Cockpits" and the "Method for Evaluating the Connectivity Capabilities of Scalable Devices in Automotive Intelligent Cockpits." He has published two academic monographs, obtained eight invention patents, and secured 12 software copyrights.

Report Title: Advanced Human-Machine Interaction Technologies for Automotive Intelligent Cockpits Empower Improved User Satisfaction

Report Introduction: 

        Based on the definitions in the Intelligent Cockpit Classification and Comprehensive Evaluation White Paper, this report introduces forward-looking technologies for automotive intelligent cockpits and how they can improve user satisfaction, focusing on four aspects: key human-machine interaction technologies, systems and components, infrastructure support, and testing and evaluation. 1. Key Human-Machine Interaction Technologies, including: perception of human status, behavior, intention, preference, emotion, and trust in natural driving conditions; real-time evaluation of performance, load, satisfaction, and safety in human-vehicle-road scenarios using edge-to-cloud integration; construction of a large cloud-based human-machine interaction model; and construction of a vehicle-side human-machine interaction calibration model based on distillation technology. 2. System and component technologies, including: the "Tuling" chassis and active suspension, AR-HUD, "one chip, multiple screens," smart seats, smart air conditioning, smart displays, smart audio, in-vehicle lighting and emotional experience, experiential and functional fragrances, vehicle-to-vehicle information interaction, and expandable devices. 3. Foundational support, including: high-performance cockpit chips, domain controllers and multi-domain collaboration, cockpit operating systems, cockpit technical architecture, cockpit digital base, artificial intelligence and large-scale model applications, and intelligent design and development tool chains. 4. Key testing and evaluation technologies, including: intelligent cockpit function and performance testing and evaluation, bench and road testing, regulatory and R&D testing, component end-of-line testing, in-vehicle real-time testing; user experience testing in semi-natural driving conditions, and real-time user experience testing using a combination of end and cloud in natural driving conditions. 5. Construction of an innovative R&D platform for intelligent cockpit human-computer interaction, empowering users with forward-looking technologies to improve user satisfaction.

About:
        Yang Zhenyu is a 15-year veteran of the automotive industry, specializing in product planning, marketing, and lifecycle management. He has led the full development and launch of several best-selling models and possesses extensive experience in user demand insights and commercialization. He has long focused on cutting-edge trends in automotive intelligence and human-computer interaction, conducting in-depth research on the application and innovation of human-intelligence interaction technologies in automotive scenarios. Combining industry practice with theoretical exploration, he has deep insights into the technological evolution and commercial value of human-intelligence interaction.

Report Title: Smart Cockpit Internationalization Challenges and the Chinese Narrative

Report Summary:  

        The globalization of smart cockpits has evolved into a dual competition of technological prowess and cultural understanding. What are the unique user profiles, technological philosophies, application characteristics, and industry chain differences across markets like China, Europe, North America, and Japan? — From Chinese speed to European order, from North American disruption to Japanese empathy. This report deeply analyzes the cultural divide underlying these differences, identifying the core conceptual differences between Eastern-style empathetic interaction and Western-style efficient control. It also proposes the essential integration of functionality, service integration, and cultural resonance required for companies expanding internationally. Ultimately, after facing the shared challenges of data sovereignty, cross-cultural trust, and algorithmic ethics, the conference proposes a globalization formula for Chinese companies: "technology as the spear, culture as the shield, and ecosystem as the king," and envisions "Made in China" taking the world by storm.

Introduction:
       Zhang Jingyu is a researcher and doctoral supervisor at the Institute of Psychology, Chinese Academy of Sciences. He currently serves on the Engineering Psychology Committee of the International Conference on Human-Computer Interaction (HCII-EPCE), an expert committee member of the User Experience Alliance, a committee member of the Complex Human Factors Division of the Chinese Ergonomics Society, a committee member of the Intelligent Cockpit Division of the China Society of Automotive Engineers, an editorial board member of CCF Transactions on Pervasive Computing and Interaction and Ergonomics in Design, and a reviewer for numerous international and domestic journals. His research interests include modeling human-computer interaction processes and behaviors in complex systems (such as driving, nuclear power, aviation, and military operations), user experience evaluation, and the design of novel human-computer interactions. He has presided over and participated in numerous national research projects (including major and key projects funded by the Natural Science Foundation of China, and key R&D projects of the Ministry of Science and Technology) and numerous corporate projects. He has published over 70 SCI/SSCI/EI-indexed papers in core human factors and human-computer interaction journals and conferences, including Human Factors, Accident Analysis and Prevention, Ergonomics, Applied Ergonomics, and the International Journal of Human Computer Studies, and has been granted 10 invention patents.

Report Title: New Challenges of Intelligence to Cockpit Human-Machine Interaction and Psychological Countermeasures

Report Introduction:

        With the advancement of intelligence, will the cockpit experience necessarily improve? How can human-machine co-driving be safe? We propose that the core issues of human-machine interaction in the intelligent era will be: how machines understand humans, how humans understand machines, and how to resolve the trust-takeover paradox. We will explore new paradigms to address these issues from new perspectives, including predicting human interaction needs, naturally expressing machine intent, and ecologically maintaining situational awareness. These ideas will facilitate the development of automated interface evaluation and optimization, new in-cabin information display systems, and vehicle-to-vehicle interaction technologies. We urge all colleagues to work together to achieve theoretical breakthroughs while building massive databases with more diverse information and engineering verification systems that require multidisciplinary collaboration.

Introduction:
       Li Hongting, PhD, is a professor and doctoral supervisor in the Department of Psychology, College of Education, Zhejiang University of Technology. His research focuses on engineering psychology, human factors engineering, user experience, and intelligent human-computer interaction. He was a technical expert in Huawei's 2012 Human-Computer Interaction Laboratory. He currently serves as Vice Chairman of the Engineering Psychology Committee of the Chinese Psychological Society. He has published over 80 papers in domestic and international academic journals and has directed three National Science Foundation projects. We have provided over 70 user experience product design and development consulting services to companies including Samsung, OPPO, Lenovo Research Institute, and China Telecom, and have co-published five textbooks and monographs.

Report Title: Human Factors Considerations for the Fluency of Voice Interaction in Smart Car Cockpits

Report Introduction:

        Fluency is the subjective experience of an individual regarding the difficulty of information processing. As voice recognition control functions have become standard features in car cockpits, the fluency of voice interaction has become a key research area of interest in the fields of human factors and user experience. This report mainly introduces the concept of fluency in human-computer interaction and the framework for fluency in smart cockpit voice interaction. It also analyzes the key human factors issues of fluency in smart cockpit voice interaction from the perspectives of perception, cognition, and feedback. Finally, it proposes a preliminary index system for fluency in smart cockpit voice interaction from the aspects of voice wake-up, voice control, voice dialogue, application control, and output feedback.

Introduction:
        Ren Wei is an expert member of the Changan Automobile AI Committee and the Director of AI Algorithms at Changan Wutong Autolink. He is primarily responsible for the independent research and development of human-computer dialogue algorithms and the development of cockpit AI interaction systems, covering speech recognition, natural language understanding, and large models. He led his team in launching the industry's first cockpit dialogue big model, the vehicle expert big model, the terminal multimodal big model, and a generative aggregation big model architecture and open access service platform. These technologies comprehensively enhance the anthropomorphism and emotionality of human-machine interaction. These technologies have been deployed in over ten Changan models, including the Gravity, Qiyuan, and Deep Blue. He presided over three key joint provincial and ministerial-level projects in Chongqing and two ITU international standards for anthropomorphic cockpit interaction. He has published six international conference papers, obtained over ten invention patents, and received the Second Prize for Scientific and Technological Progress of the China Society of Automotive Engineers and the Silver Award for Artificial Intelligence Big Model in the BRICS Industrial Innovation Competition, both provincial and ministerial level honors.

Report Title: Transformation and Practical Exploration of Cockpit Human-Machine Dialogue Driven by Big Models

Report Introduction:

        Before the advent of big models, cockpit NLU relied on a "BERT + rules" architecture, which had significant limitations: stable operation in vertical domains like vehicle control required 100,000 pieces of data, and insufficient understanding and generation capabilities for open-domain conversations, making it difficult for commercial use. Big models overcome these bottlenecks. Their powerful language capabilities enable NLU to easily process colloquial expressions, reducing the required data volume to one-tenth of the original amount. Generative capabilities significantly enhance the user experience in scenarios like mobility services and casual conversations. Leveraging agent technology and memory systems, in-vehicle conversations can handle complex tasks and meet personalized needs. Advances in large models with small parameters are driving their deployment on-board, driving the evolution of interaction towards a multimodal "language + vision" model. We are conducting research on fine-tuning efficiency, interpretability, and on-device implementation of large models.