WORKSHOPS
CHAIR
Cedric Tubert (STMicroelectronics, FR)
Daniele Perenzoni (Sony, IT)
ABSTRACT
This workshop explores cutting-edge developments in automotive vision systems, highlighting the integration of advanced cameras and photonic technologies. We present significant advancements in Automotive High Dynamic Range Imaging specifically designed for High Temperature conditions, addressing one of the industry's most challenging operational environments. The session showcases innovative wafer-scale micro-optics and meta-surfaces that are revolutionizing both imaging and illumination applications. Attendees will gain insights into the evolution of next generation CMOS image sensors for smart cars and for Driver and Occupancy Monitoring systems. The workshop also examines hardware accelerators enabling low-latency event-based vision processing, critical for real-time decision. Finally, we address the integration challenges in 'Photonics on the Road,' exploring practical hurdles and solutions for implementing these technologies in self-driving vehicles. These innovations collectively demonstrate the essential role of photonics and imaging systems in creating safer, more efficient autonomous transportation.
PROGRAM
​09:30 - 10:15
Automotive High Dynamic Range Imaging in High Temperature Conditions
Tomas Geurts (Omnivision, BE)
​The talk will cover High Dynamic Range (HDR) requirements in ADAS and In-Cabin automotive imaging applications. The importance and relevance of performance at high temperature will be explained. The talk will highlight fundamental limitations of low-light and HDR performance at elevated temperatures which is an important aspect in automotive imaging but often under-illuminated in publications.​
​
10:15 - 11:00
Past and Future of CMOS Image Sensors in Automotive Industry
Yorito Sakano (Sony Semiconductor Solutions, JP)
Business motivation is essential for the evolution of semiconductor devices. The larger the market, the faster the technology evolves. The first iPhone was born in 2007, and the back-illuminated image sensor, an epoch-making event for CMOS image sensors, was introduced in 2009. With technical breakthroughs and business motivations coming together almost simultaneously, CMOS image sensors have undergone a dramatic technological evolution over the past decade or so. Similarly, automotive CMOS image sensors have recently undergone a unique evolution in the competitive axis of high dynamic range (HDR), supported by business motivation such as the evolution of Advanced Driver-Assistance Systems (ADAS) and the efforts toward the practical application of Autonomous Driving (AD). Let me overview the recent evolution of automotive CMOS image sensors and discuss the direction of future evolution.
​​​
11:00 - 11:30
Coffee break
​
11:30 - 12:15
Wafer Scale Micro-optics and Meta-surfaces for Applications in Imaging and Illumination
Falk Eilenberger (Fraunhofer, DE)
Micro- and nanooptical systems game-changers in our ability to manipulate light. Nanooptical systems, frequently called meta-surfaces, allow to access all degrees of freedom of the optical fields, such as spectral properties, its polarization, and its phase next to its intensity, which is classically addressed in imaging systems. Nano- and microoptical systems allow to introduce massive parallelization in optical systems, breaking virtually any commonly known design rules both for imaging as well as for illumination systems. Harnessing these degrees of freedom is, however, a grand challenge in terms of design, engineering, and cost scaling. In the talk I shall highlight how wafer scale fabrication techniques can be utilized to overcome these issues, if the entire process chain from design to the final application can be tailored to the specific requirements of the optical task at hand. I shall do so by highlighting a variety of applications and projects, in which wafer scale nanooptics have played a crucial role, from optics for satellite missions all the way to illumination systems for mobility solutions.
​
12:15 - 13:00
CMOS Image Sensors for Driver and Occupancy Monitoring Solutions
Jerome Chossat and Pierre Malinge (STMicroelectronics, FR)
Automotive applications require high-performance and cost-effective sensors. Considering these constraints, we present a novel pixel architecture capable of both rolling and global shutter imaging. Utilizing a non-Bayer CFA pattern, it captures both RGB and near-infrared images. A specific ASIL pixel design ensures a comprehensive integrity check of the sensor. The latter is connected to a logic circuit through a 3D Cu-to-Cu hybrid bonding process, providing state-of-the-art on-chip data processing and interfacing. Such a sensor is ideally suited for driver monitoring systems while enabling the integration of advanced multimedia features. Indeed, on top of the pixel and readout quality requirements, CMOS mage sensors for Driver and Occupancy Monitoring solutions are bringing a lot of challenges on the digital side too. They may contain quite complex signal processing for properly dealing with various non-Bayer CFA and manage IR content, they must integrate automotive safety capabilities, must be efficiently protected against malicious attackers aimed at tampering their functionalities, and must prevent usage of counterfeit components. In addition, all this must be done under aggressive cost constraint and stringent power constraints but also be developed in conformance with road vehicles functional safety (ISO26262), and Road vehicles Cybersecurity engineering (ISO21434).
​
13:00 - 14:00
Lunch
​
14:00 - 14:45
Hardware Accelerators for Low-latency Event-based Vision
Charlotte Frenkel (TU Delft, NL)
From optical flow to high-speed particle counting, event-based cameras emerge as an enabler for low-latency vision applications. They capture temporal contrast changes as a stream of events, which are generated on a per-pixel basis and at a temporal resolution of a few microseconds. However, there is currently a lack of hardware support for event-based processing workloads that generate updated predictions within microseconds. This talk will cover emerging developments in this area, from dynamic graph neural networks to digital in-memory computing for spiking neural networks.
14:45 - 15:30
Photonics on the Road: Navigating the Integration Hurdles in Self-Driving Cars
Christoph Parl (Valeo, DE)
Valeo is at the forefront of the autonomous driving revolution, providing a comprehensive suite of sensors - cameras, RADARs, ultrasonics, microphones, and LiDARs - that enable self-driving capabilities. This keynote will explore how Valeo's technology is driving the transition from manual to fully autonomous vehicles. A key focus will be on vehicle integration: the art of seamlessly embedding these sensors into the vehicle's design. This requires balancing function-driven design, ensuring optimal sensor performance, with emotion-driven design, creating desirable and engaging vehicles. The presentation will highlight the diverse sensors required for autonomy, with a focus on LiDARs due to complexity. Crucially, we will examine the challenges and solutions surrounding sensor mounting positions. Optimal placement is vital, considering each sensor's needs, environmental factors, and cleaning requirements. Finally, we'll explore how solid-state technology can help vehicle integration to enable more compact and robust solutions for a large-scale rollout of self-driving functions.
​
15:30 - 16:00
Coffee break
​
16:00 - 16:45
Final discussion and closing​​​​
BIOSKETCHES
Jérôme Chossat
Jérôme Chossat is director of the digital architecture department and technical fellow in the Imaging Division of STMicroelectronics. In this role, he is responsible for the roadmap and execution of IP and SoC digital architectures, embedded software, image signal processing, and computer vision solutions, from early architecture to product implementation. Jérôme obtained an engineering degree in Electronics from ENSEIRB (École Nationale Supérieure d’Électronique et Informatique de Bordeaux) in 1993 and an M.Phil. degree in Microelectronics from Bordeaux University in the same year. He has been working for STMicroelectronics since 1996, when he joined a team designing image processors for televisions and webcams. He has been involved since the early days of mobile phone imaging, developing families of image signal processors for mobile phone applications and directly contributing to very high-volume mobile phone and camera systems. He has also been involved for 15 years in the development of several generations of automotive image sensors.
​​
Charlotte Frenkel
Charlotte Frenkel is an Assistant Professor at Delft University of Technology, The Netherlands, since July 2022, and a Visiting Faculty Researcher at Google since October 2024. She received her Ph.D. from Université Catholique de Louvain in 2020 and was a post-doctoral researcher at the Institute of Neuroinformatics, UZH and ETH Zurich, Switzerland. Her research aims at bridging bio-inspired and engineering-driven design approaches toward neuro-inspired AI systems, with a focus on digital neuromorphic processor design, embedded machine learning, and on-device learning. Dr. Frenkel serves or has served in the technical program committee of various conferences such as DATE, ESSERC, and NICE, co-leads the NeuroBench initiative for benchmarks in neuromorphic computing, and is an associate editor for the IEEE Transactions on Biomedical Circuits and Systems.
​
Tomas Geurts
Tomas Geurts received the M.S. degree in electrical engineering from the Katholieke Universiteit Leuven, Belgium, in 1999. After graduating, Tomas worked at AnSem where he focused on RF and SerDes IC design. In 2004 he joined FillFactory, which was later acquired by onSemi, where he held various leadership roles in the image sensor division until he took on a role as CTO for the industrial sensing division. At onSemi, Tomas played a key role bringing LOFIC technology to maturity for cinematography and automotive applications. In May 2021 Tomas joined OMNIVISION where he is currently a Sr. Director, heading up R&D for OMNIVISION in Europe, and a member of the office of the CTO.
​​
Falk Eilenberger
Falk Eilenberger has studied Physics in Jena. After graduation he received an Endevaour Europe award of the Australian government and worked in the field of nonlinear photonics at the University of Sydney. He returned to Jena’s Institute for Applied Physics and received his PhD on research in Spatiotemporal Nonlinear Photonics in 2014. He then held a position as research coordinator for the Fraunhofer-Institute of Applied Optics and Precision Engineering IOF before transition to a position of junior research group leader on Photonics in 2D-materials in 2018. He transitions to become IOF’s department head for Micro- and Nanophotonics in 2022. Falk Eilenberger is an expert in nanophotonic, waveoptics, nonlinear photonics, and quantum optics and holds the Thuringian Innovation Prize in 2024 for his efforts to combine novel materials with nanooptical systems.
​​
Pierre Malinge
Pierre Malinge joined STMicroelectronics France in 2013, where he began working on imager pixel design. At STMicroelectronics, he has worked on many technologies and contributed significantly to establishing the foundation of the global shutter product family. Before joining STMicroelectronics, Pierre gained valuable experience in advanced memory design and SRAM bitcell development at leading semiconductor companies. He received his M.Sc. degree in Electronics in 2002 from ESEO, Angers, France, and his Ph.D. degree in 2005 from INSA de Lyon, France.
​
Christoph Parl
Since 2018, Dr. Christoph Parl has been developing LiDAR sensors for Valeo. Here he became a system design leader and senior expert, bridging the gap between industry and academia. Prior to this, he led a research group at the Werner Siemens Imaging Center. By education he is an electrical engineer who drifted towards physics during PhD, which he received in 2016 from the University of Tübingen, Germany. Throughout his 20-years professional career, Christoph has been active in medical imaging, high-energy physics, agriculture, and avionics. Across these diverse fields, he has consistently focused on improving low-light sensing, an absorbing topic that has stayed with him to this day.
​
Yorito Sakano
Yorito Sakano received the M.S. degrees in applied chemistry from the Graduate School of Engineering, Tohoku University, Sendai, Japan in 1999. In 2000, he joined Seiko Epson Corporation, Sakata, Japan for the development of Vth modulation image sensors. In 2007, he moved to Sony Corporation, Atsugi, Japan for the development of back-illuminated CMOS image sensors and high-dynamic-range global-shutter CMOS image sensors. He is now involved with the development of advanced CMOS image sensors for automotive applications in Sony Semiconductor Solutions Corporation.