At the 2026 Aggregated Intelligence Industry Development Conference held on September 9 in Wuhan's Optics Valley, jointly organized by the Aggregated Intelligence Industry Innovation Center and Hubei Science and Technology Investment, Zhang Yongwei, chairman of the Auto 100 Research Institute, delivered a keynote speech highlighting the distinct challenges facing two emerging sectors. The automobile industry is grappling with intensifying competition and compressed profit margins, making it imperative to explore new growth avenues to escape its current predicament. Conversely, the embodied intelligence industry's primary hurdle lies in industrial-scale deployment, particularly in reducing costs and transitioning technology from laboratory settings and demonstration prototypes to real-world applications.

Zhang stressed that despite these differing pain points, the two sectors should pursue integrated development across several key areas. The first priority is accelerating the creation of a unified AI technology foundation. Looking back at the automotive industry's trajectory, it progressed from early small-scale models to end-to-end solutions proposed in 2023, and by last year, industry consensus had shifted toward large models as the core. Without high-quality foundational models, it becomes extremely difficult to support the deployment of two highly complex autonomous systems in intelligent driving and smart cabins, which is why the industry has increasingly focused on large model development. Looking ahead, a hybrid model approach is emerging, enabling gradual implementation and iterative value creation along the way. Although the model architecture for robotics is more complex, the technological evolution path is fundamentally similar, confirming that both sectors share a common direction: building a unified AI foundation that can be applied and reused across different domains.

Second, both industries should jointly develop hardware infrastructure across storage, computing, connectivity, and sensing. On storage, both autos and robotics faced significant challenges from rising memory prices over the past year, and without robust storage chip capabilities, supporting AI, intelligent vehicles, and embodied systems will remain difficult. On computing power and processing, particularly high-performance AI chips, this represents the most innovative area within the aggregated intelligence space, where domestic alternatives are increasingly being adopted in vehicles despite previous bottlenecks. Third, connectivity is gaining growing attention with vast innovation potential, especially next-generation technologies such as optical communications, wireless networks, and satellite communications, all of which will greatly expand the functional capabilities of intelligent terminals. Fourth, sensing is essential across all platforms, including aerial, land-based, and ground vehicles, all requiring substantial sensor equipment deployments that push forward innovations in new materials and advanced manufacturing processes. These four pillars constitute the most fundamental layer of intelligent and AI hardware.

The third recommendation focuses on streamlining the mechanical industry chain spanning materials, components, and actuators. Robotics allocates 50 to 70 percent of its costs to mechanical structures, and the automotive sector carries similar burdens while also incorporating battery systems. The various actuator-based foundational products equipped with storage, computing, connectivity, and sensing capabilities serve as the core components. Establishing a complete chain from materials and parts to actuators represents the optimal fusion of the two industrial supply chains. By promoting the integration and migration of automotive supply chains into the embodied intelligence ecosystem, companies can simultaneously deepen their presence in the automotive sector while expanding into embodied intelligence, continuously achieving bidirectional technology transfer and unlocking entirely new market opportunities for the automotive industry.

Fourth, leveraging scenario resources is crucial beyond the migration of AI hardware and mechanical components. Automotive manufacturing plants themselves present highly suitable application scenarios for embodied intelligence, particularly in high-risk and physically demanding operations such as vehicle final assembly, quality inspection, battery disassembly lines, and hazardous factory logistics, spanning both vehicle and component production processes. Using real-world scenarios to drive embodied technology deployment can forge a distinctive pathway for the growth of China's emerging industries.

Fifth, aggregated intelligence should both draw lessons from and transcend the automotive industry's development experience. The automotive sector followed a trajectory of domestic market capture first, followed by overseas expansion. However, based on recent international research, Zhang argued that the embodied economy should adopt a globally-minded approach from its inception, balancing both domestic and international markets from the start. Overseas markets may even become early adopters for numerous new technologies and application scenarios, creating fresh opportunities for growth.

Finally, building a comprehensive standards, testing, and certification framework is essential for the advancement of aggregated intelligence industries. The automotive sector established vehicle-grade standards from its very beginning, prioritizing human safety as the fundamental baseline and developing a mature, comprehensive system with the highest certification requirements. Leveraging this established vehicle-grade framework to collaboratively establish standards, testing, and certification norms for embodied intelligence represents another critical lever for integrating the automotive and embodied intelligence sectors.