Agenda

Kazuko Nishimura | Senior Researcher of R&D Center, Sony Semiconductor Solutions Corporation

While image sensors seem to be well established and stable from a marketing viewpoint, there is much going on in the technology arena that is differentiating between design and manufacturing. These include a shift in the technology direction from pixel scaling to adding functionality, from organic microlenses and colour filters to metastructures, from the visible spectrum to extensions to shorter and longer wavelengths, from computational image sensor processing to the quest to incorporate neural networks where image sensors live at the "edge of the cloud". In this talk these areas will be explored (or exposed) by reaching into the reverse engineering imagery and data collected by TechInsights from the latest image sensor products. The goal will be to place them into context of trends past and to provide a "crystal ball" view of upcoming decision points for the technology.
Dan McGrath | Senior Technical Fellow, Image Sensor, Tech Insights

The CMOS image sensor (CIS) industry has entered a renewed growth phase after a challenging 2023. In 2024, revenues rose by 6.4% year-on-year to reach $23.2 billion, driven by the smartphone rebound, expanding automotive and security applications. The market recovery has been supported by domestic demand in China for smartphones and consumer products, while rising competition in areas such as automotive has added further momentum to the industry. On the technology side, the industry is moving beyond pixel scaling toward performance, integration, and versatility. Stacked architectures now account for most production, and progress in 3D sensing, event-based imaging, and metasurface optics is enabling more compact and capable sensors for diverse environments. The presentation will cover CIS market and technology trends, ecosystem dynamics, and new application opportunities.
Anas Chalak | Market & Technology Analyst – Imaging, Yole Group

Sara Pellegrini | Advanced Photonics Pixel Architect, STMicroelectronics

Stitching is a technique used to make large area image sensors, but stitching comes with a price. We will discuss: • A general overview of the history of stitching,
• Stitching with 1 reticle or 2 reticles per layer
• Why is stitching so expensive?
• Why are the fabs hesitating to allow stitching in their (CIS) processes?
• What about stitching in combination with stacking?
Albert Theuwissen | Founder, Harvest Imaging

A number of compelling technologies are competing to establish the low-cost SWIR image sensor market with decent volumes: colloidal quantum dots, organics, Ge-on-Si and low-cost InGaAs. We will review the maturity of these image sensor technologies in terms of performance, reliability and scale-up prospects. We will discuss the markets that are being targeted with their challenges, which will provide a perspective on the path ahead for these technologies.
Andras Pattantyus-Abraham | Founder, Best QD Solutions LLC

For more than fifty years, the basic method behind colour imaging has changed very little. Nearly all CMOS image sensors still rely on the Bayer colour filter array, a grid of red, green, and blue filters that allows silicon sensors to capture colour. In this mosaic arrangement, each pixel is covered by a single red, green, or blue filter, meaning only about one-third of the incoming light contributes to the pixel’s signal. Consequently, close to seventy percent of photons are lost before they reach the photodetector. This loss reduces signal-to-noise ratio (SNR), a long-standing limitation for users of smartphones and compact cameras. A new phase in imaging is now emerging. Nanophotonic colour splitting technology replaces filtering with a system that guides light through sub-diffraction-limited waveguides. This approach removes the inefficiencies of Bayer-based sensors and opens the way for ultra-compact, high-resolution, and light-efficient cameras in smartphones, XR devices, industrial inspection systems, and medical diagnostics. The presentation will cover this innovation, which replaces the Bayer filter with a nanophotonic waveguide layer that splits light by wavelength and directs it precisely to each pixel. Instead of absorbing unused wavelengths, the design employs vertical waveguides to separate colours, channelling photons with minimal loss and maximum resolution.
Jeroen Hoet | CEO & Co-founder , eyeo

As pixel pitches shrink and optics get thinner, image quality is increasingly constrained by shot-noise-limited capture, pixel cross-talk, microlens effects, diffraction/aberrations, and the complexities of Quad/Hex and other modern CFAs. This talk explores how an end-to-end Neural ISP can replace or augment multiple classical pipeline stages to restore detail and color fidelity directly from RAW, while remaining compatible with camera engineering validation. We’ll share practical lessons on robustness, evaluation, and edge deployment realities (latency, memory, quantization, and HDR modes), and close with a forward look at sensor–optics–ISP co-design and the tradeoffs between hardware complexity, cost, and computation.
Tom Bishop, Ph.D., Founder, CTO, Glass Imaging

Markus Cappellaro | Product Manager and Principal, Carl Zeiss Microscopy Technology Center Munich

• AR glasses & AI glasses – system architecture, sensor requirements, pixel requirements, compute requirements, camera/imaging requirements
• ISP & SoC: needs for the future
• How are IMUs, cameras & sensors used for AI?
• How does depth sensing fit into the future of AI & AR?
Harish Venkataraman | Principal Engineer, Camera & Depth Systems, Reality Labs, Meta Inc.

Vision systems are currently driving a significant technological shift, enabling a rapidly expanding and still emerging set of applications across domains such as humanoid robots, delivery drones, biometric payment, access control, and smart building automation. This ongoing transformation is challenging traditional design paradigms, demanding higher levels of integration, stringent power efficiency, and optimized processing architectures. This presentation examines the technical synergy between advanced on-chip image sensor processing and SoC computation as a key approach to meet these evolving requirements. Key innovations include dual shutter pixel technology that integrates both global and rolling shutter operation, providing flexible frame capture strategies to mitigate motion artifacts while supporting high dynamic range imaging. The sensor architecture incorporates on-chip RGB-IR separation combined with integrated bayerization, IR depollution, and smart upscale algorithms, enabling simultaneous full-resolution RGB and near-infrared imaging optimized for a wide range of illumination conditions, including HDR outdoor scenarios. By offloading computational tasks such as colour separation and preprocessing into the sensor, the data throughput and processing load on the SoC are significantly reduced, enabling the use of lower-power microcontrollers devices without compromising system performance. The presentation will also highlight typical implementation case studies across representative embedded vision scenarios, demonstrating how this sensor-SoC synergy supports efficient, scalable, and adaptable machine vision solutions.
Marie-Charlotte Leclerc | Product Marketing, Industrial, Security & IoT Imaging, STMicroelectronics

In recent years, Industrial automation is advancing at a rapid pace. For this, a wide variety of sensing technologies are becoming increasingly important. Learn about Sony's contribution to solving such challenges by utilising its sensing technology including global shutter sensors, SWIR, UV, and EVS.
Roberto Buttaci | Head of Industrial Sensors, Sony Semiconductor Solutions Europe

Vladimir Koifman | Chief Technology Officer, Analog Value

Sensing is rapidly evolving, with autonomous systems demanding increased sensor resolutions, faster refresh rates, and wider colour depths. All this leads to a rise in the bandwidth requirements of connectivity technology. MIPI A-PHY is the first industry standard to deliver multi-gigabit, long-reach connectivity purpose-built for automotive and used for machine vision, endoscopies, and professional audio-video. This session explores how A-PHY enables direct, low-latency connections from cameras, radar, and LiDAR sensors, natively extending the widely used MIPI CSI-2 protocol. We’ll discuss key benefits for OEMs: increased resilience to electromagnetic interferences (EMI), future-proof scalability, interoperability that comes from a multi-vendor ecosystem. Lastly, we'll review how the A-PHY standard enables innovations in sensor technology by allowing vendors to directly integrate A-PHY into the sensor, leading to smaller, lower power, lower cost vision systems. MIPI A-PHY is the first industry standard to deliver multi-gigabit, long-reach connectivity -Purpose-built for automotive and used for machine vision, endoscopies, and professional audio-video -Allows vendors to directly integrate A-PHY into the sensor, leading to smaller, lower power, lower cost vision systems
Arno Distel | Field Application Engineer, Valens Semiconductor

Image sensors are proliferating in key application areas such as automotive ADAS and user experience, industrial automation and machine vision, and security/biometric identification. Traditionally, ceramic packages have been used to house image sensors, as these give good optical, thermal and reliability performance. However, such ceramic packages come at a high cost, thus packaging structures migrated to laminate-based packages such as iBGA to provide more attractive price points and help increase the adoption of image sensors in new application spaces. As higher resolution sensors are now being demanded by the market, the image sensor die size is increasing, resulting in the need for larger package sizes which start to approach reliability and mass production efficiency limits of traditional iBGA packages. A new large body molded image sensor has been developed to be able to address larger image sensor package sizes with better reliability performance, whilst also offering productivity improvements in high volume manufacturing operations.
Alastair Attard | Director of Business Development - MEMS & Sensing Applications, UTAC Group

• What if supply from Taiwan breaks away tomorrow? - Is it an existential risk to sensor and camera OEMs? What are the alternatives? What initiatives are underway to reduce dependency?
• What if China shuts down the supply of rare earths? - Is it an existential risk to sensor and camera OEMs? What are the alternatives? What initiatives are underway to reduce dependency?
• What can be done to optimise the cost of manufacturing? - Is the impact of raw materials significant? Special focus on SWIR sensors.
• How can sensor and camera makers support foundries in maximizing utilisation / cost reduction? Can "out-of-blemish-spec" sensors be put to use?
• How loyal are sensor OEMs to their foundries? Is there enough competition between foundries?
 
Dr. Ronald Müller | Associate Consultant, Smithers & CEO, Vision Markets

At the end of day one of the conference, come together for unrivalled networking with your fellow industry peers over drinks and canapes. The drinks reception will run until 8:30pm, leaving plenty of time for discussion and the opportunity to then extend you evening in the city for dinner.