As the world's first supplier of 100GigE cameras, Emergent Vision Technologies (EVT) has been deeply involved in the high-speed machine vision field for many years. Qin Yongsong, Regional Manager for China, believes the wave of large AI models presents a significant development opportunity for vision technology. "This enables us to enhance performance and increase added value, helping us avoid getting caught in the vortex of low-cost competition," he stated. From embodied intelligence to new energy applications, EVT is supporting industry intelligent upgrades with its technological advantage of achieving synchronization precision down to 1 microsecond.

Interviewee: Qin Yongsong, Regional Manager for China, Emergent Vision Technologies

Q1: Could you briefly introduce your company's historical background and its core competitive advantages in the machine vision field?

Qin Yongsong: Founded in Vancouver, Canada, Emergent Vision Technologies holds the distinction of being the world's first supplier of 10GigE, 25GigE, 50GigE, and 100GigE cameras. We have consistently been dedicated to providing innovative and reliable high-speed camera solutions for cutting-edge applications demanding high speed and high data throughput.

To date, we have deployed GigE Vision cameras into numerous innovative applications. These include machine vision inspection, motion analysis technology, broadcasting, volumetric capture for virtual worlds, intelligent transportation systems, virtual reality (VR), augmented reality (AR), and life sciences. Our products excel not only in superior image quality but also offer strong competitive advantages in the high-performance transmission of image data and the efficient operation of vision systems.

Q2: Which latest products or cutting-edge technologies did your company prominently showcase at the VisionChina 2025 (Beijing) Machine Vision Exhibition?

Qin Yongsong: Recognizing Beijing as a key hub for research and academic studies, we leveraged this significant industry platform not only to exhibit our mature 100GigE cameras, our self-developed NICs with GPU Direct functionality, and our 3D point cloud generation application software tailored for motion capture, but also to highlight our newly developed EROS series.

This series boasts the industry's smallest form factor and lowest power consumption for 10G high-performance cameras. It offers dual interface options (RJ45 and optical fiber) and supports multi-tasking functionality. A single PC can interface with dozens of cameras simultaneously while maintaining synchronization stability within 1 microsecond.

Q3: In the pursuit of higher precision, which accuracy requirements encountered in practical scenarios do you consider excessive or pseudo-demands?

Qin Yongsong: We have indeed encountered this issue in practical applications. Some customers request 16-bit visual data. Based on our experience, 16-bit resolution is typically excessive.

In reality, 8-bit data generally suffices for the majority of applications. Our cameras, capable of providing very high-quality data at 10-bit or 12-bit, can fully meet customer requirements by outputting 8-bit data. The demand for 16-bit often represents a case of over-specification with poor cost-effectiveness.

Q4: Beyond traditional industrial inspection, machine vision is now empowering increasingly diverse scenarios. Which emerging fields do you view with the most optimism?

Qin Yongsong: We are particularly optimistic about two emerging sectors: New Energy and Biopharmaceuticals. In the New Energy sector, China's rapid development, having met widespread application needs, is now driving demand for higher-level technological upgrades. This trend imposes progressively higher requirements on vision technology.

Within Biopharmaceuticals, customers are now looking beyond simple image capture. Requirements for backend data transmission, multi-camera applications, and related scenarios are becoming increasingly sophisticated. This aligns well with the key development directions we prioritize.

Q5: Are there specific collaborative cases or strategic initiatives currently underway within these fields?

Qin Yongsong: Our product and technology roadmap offers distinct differentiation. Previously within the industrial sector, engagements were primarily customer-initiated. We are now proactively engaging with customers, especially within the New Energy sector – encompassing battery manufacturing, wind energy, and even nuclear energy applications.

We observe a positive shift: customers, evolving from initial unfamiliarity, can now articulate precise technical requirements. This marks an excellent starting point. We are also reallocating resources towards promotion, witnessing growing adoption in Biopharmaceutical applications like genetic testing and cell analysis. These represent our current key focus areas within the domestic Chinese market.

Q6: With the rise of the embodied intelligence trend, what new demands and characteristics do you believe this imposes on machine vision systems? Could it catalyze the emergence of new production lines?

Qin Yongsong: Embodied intelligence is indeed a significant trend, with humanoid robots being a prominent current application. This strongly propels our vision technology forward, as data capture at the front end – whether for humanoid or other robots – constitutes the essential first step.

Compared to traditional sensor solutions, embodied intelligence leans more towards simulating human perception, utilizing "eyes" for vision. This imposes stricter demands on vision systems: latency must be minimized, synchronization between multiple cameras must be enhanced (e.g., mimicking human binocular vision), and computational resource consumption needs reduction. These requirements are indeed becoming increasingly stringent, but they simultaneously drive our technological investment and R&D improvements in this domain.

Q7: What specific technological updates has your company implemented in response to embodied intelligence requirements?

Qin Yongsong: To address the demands of embodied intelligence, we have executed targeted product upgrades primarily manifested in three critical aspects: ensuring zero frame loss during data transmission; guaranteeing extremely low latency with minimal jitter; and maintaining multi-camera synchronization precision within 1 microsecond. Collectively, these enhancements establish a more reliable data foundation for backend intelligent processing in robotic systems, thereby reducing the potential for errors.

Q8: Regarding the shift from product sales to service provision, what is your perspective on the "Vision as a Service" (VaaS) business model? What is the current level of customer acceptance?

Qin Yongsong: This evolution genuinely reflects changing industry dynamics. With one to two decades of experience in this sector, I initially observed customers favoring comprehensive solutions due to limited understanding of vision technology. Subsequently, advancements in China's industrial capabilities enabled engineers to master these technologies, leading them to primarily request individual components.

Presently, amidst further industrial upgrading, engineers—while technically proficient—lack dedicated bandwidth for vision system integration, necessitating focus on other product development priorities. Consequently, demand has shifted back toward turnkey solutions where we provide end-to-end services. Concurrently, escalating requirements for technological upgrades and value-added features elevate expectations to another tier, mandating increasingly holistic service frameworks.

Q9: Given the growing customer emphasis on Total Cost of Ownership (TCO), how do you propose redefining product value?

Qin Yongsong: My professional trajectory—progressing from components to solutions, integrated projects, and now ecosystem offerings—has crystallized this concept. Historically, customers prioritized product-centric cost and value. Contemporary clients increasingly evaluate holistic packaged pricing encompassing lifecycle maintenance expenditures and ancillary costs incurred by integrating our products into their equipment or projects.

For instance: whereas traditional camera systems might require 4-5 servers, our solution supports 40-50 cameras per server. This exemplifies the TCO philosophy and aligns with our strategic direction. Customers are demonstrably recognizing the criticality of this paradigm.

Q10: Under the wave of large AI models, will traditional machine vision algorithms face disruption or empowerment?

Qin Yongsong: AI unequivocally empowers industrial vision—even catalyzing advancement within our domain. It represents an inflection point for industry resurgence or accelerated growth.

AI, emphasizing artificial intelligence and human cognition emulation (paralleling embodied intelligence), inherently prioritizes data acquisition as its foundational element. Rapid progress in algorithm development by abundant talent pools intensifies demands on front-end hardware. Consequently, pressures from cost-driven sectors like 3C electronics on traditional vision are alleviated. The AI surge enables performance enhancement and value addition, allowing avoidance of low-cost competition vortices. Thus, this presents a significant strategic opportunity.

Q11: EVT exhibits in Beijing, Shanghai, and Shenzhen. What distinctive characteristics differentiate customers across these regions?

Qin Yongsong:These three regions exhibit distinct market profiles:

Beijing attracts primarily R&D-intensive clients with stronger emphasis on innovation capabilities, demonstrating relatively lower cost sensitivity despite moderate market volume.Guangdong serves predominantly end-user applications with consistently high-volume demand, though characterized by stringent cost-control requirements.Shanghai functions as a comprehensive ecosystem attracting diverse clients across geographical and functional spectra – including both R&D entities and end-users – resulting in the largest market scale due to its integrative nature.

This structural differentiation aligns with Beijing's strategic positioning as an innovation hub, where the expanding technical workforce continues to drive specialized demand growth independent of volume-based constraints.

Q12: How would you evaluate the VisionChina 2025 (Beijing) exhibition?

Qin Yongsong: The exhibition exceeded expectations. Attendees demonstrated exceptional enthusiasm and technical proficiency. Coupled with our distinctly differentiated product showcase, this attracted substantial engagement from specialized clients. We are highly satisfied—the event effectively demonstrated our technological capabilities while providing deep market insights. We will unequivocally continue participating in the Beijing exhibition next year.