• 133 attendees from 23 countries, 80 B2B networking meetings
• Conference highlights AI and Edge computing
• European players in good starting position

Barcelona/Stuttgart, 04 November, 2019. Organized by the EMVA and Messe Stuttgart, the Embedded Vision Europe 2019 (eVe) conference attracted over 130 international conference attendees from 23 countries from 24 – 25 October at the ICS Stuttgart. They represented the entire value chain from deep learning and AI developers, machine vision players to users of embedded vision systems and even stakeholders from the financial sector. The presentation program included the full spectrum of the embedded vision industry, covering topics such as new hardware compute platforms, embedded vision standards and APIs, specific approaches to optimizing neural networks, as well as real-world examples of the deployment of vision based embedded AI systems. It was framed by a highly frequented table-top exhibition. Cooperation and exchange between the attendees was supported by close to 80 individual booked B2B meetings between the attendees during the conference breaks.

Dynamic area for innovation

EMVA Board of Directors member, Dr. Chris Yates moderated the conference and in his opening remarks stated that “Embedded vision is one of the most dynamic and creative areas for innovation in our industry”. He noted that the importance of embedded vision is demonstrated by the remarkable fact that there are almost certainly more embedded vision systems in use today than humans on the planet. “These systems are transforming our factories, hospitals, transport, and living spaces, by providing machines with greater cognitive ability through vision”, Yates said.

Artificial Intelligence (AI) as a dominant topic  

Artificial Intelligence and with it Deep Learning proved to be a dominant topic in many of the speeches. For good reasons, as David Austin from Intel in his opening keynote showed: He stated that AI is transforming Industry 4.0, which is introduced into machine vision industry by the OPC-UA standard, and all the industrial verticals. Thus it soon will be everywhere, impacting manufacturing, energy, logistics and building. With regards to the significance of industrial production within the economy Austin cited a consultant’s estimate that the economic impact of factory IoT applications will reach $1.2 trillion by 2025. Since deep learning models can only describe what they’ve been trained to see he presented three techniques how to come to practical and flexible AI based solutions. His key message was that AI is still in the early adoption stage and it is time to get on board now.

While it was made clear that the continued development of embedded neural networks is finding widespread adoption and providing real benefits to end users, also topics such as real-world deployment and maintenance were addressed by several presenters, indicating a growing maturity of the field as the focus moves beyond pure technical performance of a AI based application. Vassilis Tsagaris, CEO of Irida Labs put a focus on this aspect in his talks and further explained his point: „The way AI and Deep Learning models are trained and deployed for Embedded Vision needs to move beyond the technology hype and be focused on customer needs. What is currently still missing in the technical world of developing AI is a holistic problem solving approach, including not only training deep learning models, but also validation, inference at the edge devices and continuous update mechanisms. For this purpose, we need to look beyond the machine learning models and deploy AI-based embedded vision in real world products.”

Computing at the Edge

Many of the presentations given addressed edge computing, wherein data which is not sent to a centralized place or database but processed at the source, and thus intelligence is moving into systems as one key characteristic of embedded systems. One example was Jagan Ayyaswami from Micron Technology who presented his thoughts on how to choose an embedded architecture and to answer the question ‘How do we keep the data as close as possible to the image capture’. Ratislav Struharik from IDS pointed out that edge computing requires low latency, time-sensitivity, and task-specific processing at relatively low cost in power, price and size. The same would hold true for AI on the Edge.

 European embedded vision strengths made visible

Struharik also made an important point as a European representative presenting a radical new solution at IDS by adapting the app principle to industrial image processing though so-called vision apps. Another truly European contribution was the lecture given by Dr. Konstantin Schauwecker from Nerian Vision showing that stereo vision is a compelling technology for 3D depth sensing and FPGAs make stereo vision usable for embedded use cases. His presented work also proves that vivid interaction between academic research and start-up entrepreneurship in Europe can lead to prosperous products which have all chances to find their market place.

A practical speech was given by Gion-Pitschen Gross from Allied Vision, another European player, who focused on the requirements, components and solutions to set up an embedded vision system. He pointed out that the overall system design of an embedded vision system is highly dependent on the use case and many different skillsets are involved. In his conclusion he noted that despite the attractiveness of embedded vision systems, on many occasions the design costs of an industrial embedded vision system are still higher than of a PC-based system put together mainly from off-the shelf products; and their integration efforts are often underestimated which quickly can eat up the predicted ROI.

Andrea Dunbar from the Swiss research center CSEM impressed the audience with a project that achieved to develop a coin-size 3mm thick fully autonomous camera which is deployable like a sticker. It uses an ultra-low power image sensor consuming less than 1mW that is even able to perform custom image processing such as motion detection.

Computational Neural Networks are here to stay

In his closing remarks Dr. Chris Yates stated that AI and computational neural networks have become a dominant topic which will be here to stay. While technological limits have been driven further and further recently the current challenge would be to put a focus on ease of deployment and thus abstracting the complexity of the products that find their way into machine vision applications. For this, he pointed out; further cooperation amongst small and big players would be needed.

About EMVA:

Founded in May 2003 in Barcelona, the European Machine Vision Association currently has about 120+ members representing more than 20 nations. Its aim is to promote the development and use of machine vision technology and to support the interests of its members – machine vision companies, research institutions and national machine vision associations. The main fields of work of EMVA are: standardization, statistics, the annual EMVA Business Conference and other networking events, European research funding, public relations and marketing. To find out more visit the web site www.emva.org.

User Level for base knowledge about the standard and expert level to perform own EMVA 1288 measurements

Barcelona, 29 October, 2019. For almost ten years release 3 of the EMVA 1288 standard has been available for an objective and application-oriented characterization of industrial cameras and is being used worldwide. Today the rapid and error-free development of new cameras is no longer conceivable without accompanying EMVA 1288 measurement. OEM customers can use the EMVA 1288 parameters to realistically and precisely define the requirements for customer-specific image sensor development. Distributors, system developers and end customers also use EMVA 1288 to select the best camera based on the requirements of an application.

New EMVA 1288 Certification Program

The way the EMVA 1288 standard is structured makes its application as simple as possible. However, image sensors and industrial cameras are complex systems. Their best possible use is not managable without appropriate knowledge. EMVA therefore offers, in cooperation with member companies, training courses on the practical application of the standard, which can be combined with certification examinations on two levels. These two levels are:

The User Level includes the necessary knowledge to select the best possible camera(s) for a given application based on EMVA 1288 data sheets. The target group of this certification is employees from marketing and sales as well as all technically oriented employees who need basic knowledge about the EMVA 1288 standard. A total of four free webinars in English provide the knowledge required for certification on the User Level:

1. Why is an image sensor not perfect? Or – what effects degrade the signal of a camera?
2. Basic parameters of the EMVA 1288 standard: What do they mean and how are they measured?
3. What you should know about your application to select the best camera for it.
4.  How to perform camera comparison in practice using the EMVA 1288 summary datasheet?

Certification exams at the User Level will be offered next year during the two trade fairs Control (Stuttgart) in May 2020 and Vision (Stuttgart) in November 2020. Details will follow soon. The certificate proves that the owner understands the important parameters of an image processing application and has basic knowledge about image sensors, cameras and the meaning of the EMVA 1288 parameters.

The Expert Level is designed for all those who want to develop cameras or perform EMVA 1288 measurements themselves. This requires practical experience with EMVA 1288 measurements. For this, in cooperation with two member companies and for a participation fee EMVA offers two-day training courses with practical demonstrations. At the end of the second day there will be the opportunity to take the written certification exam as an EMVA 1288 expert. The certificate shows that the holder has all the necessary knowledge to successfully carry out EMVA 1288 measurements himself and to interpret their results correctly, in particular if they deviate from the ideal behavior.

The next trainings for the EMVA 1288 Expert Level are:

• 4.- 5. December 2019, AEON, Hanau
• 13.- 14. May 2020, Framos, Munich   (+++Attention! Online Training May 27th – 28th, 2020+++)

 Great benefits of certification at very manageable costs

Applicants who pass the certification test of the User Level or the Expert Level receive a corresponding certificate. With this certification employees as well as companies can objectively prove and promote their EMVA 1288 knowledge. Given the widespread use of the EMVA 1288 standard, the certification costs of 200 Euros per Level for EMVA members and 250 Euros for non-members are manageable and equally useful investment.

About EMVA:

Founded in May 2003 in Barcelona, the European Machine Vision Association currently has about 120+ members representing more than 20 nations. Its aim is to promote the development and use of machine vision technology and to support the interests of its members – machine vision companies, research institutions and national machine vision associations. The main fields of work of EMVA are: standardization, statistics, the annual EMVA Business Conference and other networking events, European research funding, public relations and marketing. To find out more visit the web site www.emva.org.

Common Press Dossier from EMVA and Lakesight Technologies

Stresa, Lago Maggiore; Barcelona; Munich 18 October, 2019. Machine vision engineers from all over the world have gathered from 07-11 October in Stresa/Italy at the shore of Lago Maggiore for the International Vision Standards Meeting (IVSM) Fall 2019. This meeting takes place twice a year under the global G3 initiative which is supported by the machine vision associations AIA, CMVU, EMVA, JIIA and VDMA. Hosting association in Stresa was the European Machine Vision Association (EMVA) together with the corporate sponsor Lakesight Technologies. Birger Engineering and iMGAGE S were co-sponsors of the event.

IVSM agenda Standard Working Groups; Plugfest; and FSF

Each machine vision standard has its own working group which works continuously on the further development of the standard. Standard group members are engineers from machine vision companies and machine vision scientists. They meet personally twice a year during the IVSM. At the fall 2019 IVSM in Stresa meetings took place from the standard working groups of CameraLink, CameraLink HS, CoaXPress, emVISION, GenICam, OOCI, and OPC-UA Vision.

All standard working groups are open for new members who are interested to contribute with their experience.

One highlight of each International Vision Standard Meeting is the so-called Plugfest. During this event it is possible to test any exciting or new device with most worldwide available host-applications for interoperability. As the Plugfest attendance is restricted to people from the standard working groups and/or programmers from the involved companies most problems can be discussed and sometimes even solved short-term without the usual barriers where these engineers meet such as during trade shows or in a customer usage situation. The Plugfest has its roots in the development of the GenICam standard and nowadays includes practically all machine vision interface standards present during an IVSM.

Last but not least, the Future Standards Forum (FSF) in its General Meeting session gives an update on all current machine vision standards as well as an outlook on the date and place of the forthcoming International Vision Standard Meetings which are always presented by the respective hosting associations.

EMVA and Lakesight Technologies underline importance of standardization

Even after many years in the standardization community EMVA Standards Manager Werner Feith was once again impressed about the growing number of participants and the high-level output of the meeting. “Looking at the manpower and effort put into standardization makes clear that machine vision standardization cannot be taken for granted. A lot of industries look with envy at the level of commonly developed standards in the machine vision industry, where even engineers from competing companies jointly work together in the various standard working groups”, says Feith and adds: “As a machine vision association being part of the global G3 initiative the support of standardization activities is written in EMVA’s DNA. One cannot overestimate the value standardization has brought to our industry by means of facilitating machine vision applications and thus customer adoption of the technology; but also avoiding double developments.”

Lakesight Technologies as first time corporate IVSM sponsor in Stresa is equally satisfied after one week of intensive discussions about machine vision standardization. “Organizing and hosting the IVSM in Stresa was an experience of its own kind. We are glad that Lakesight Technologies was given this opportunity. Witnessing the serious progress that was made in the various standard working groups by technicians from all parts of the world at this fall gathering we realized it was the right decision”, says Lakesight’s CEO Business Unit Vision Solutions Martin Hund and adds: “Our engagement was also a statement that despite the current phase of ongoing market consolidation progress in standardization can only continue through further cooperation. There is absolutely no reason to question the intensive collaboration taking place during the International Vision Standard Meetings.”

The EMVA hosts several machine vision standards, namely the two widely established standards GenICam and EMVA 1288. In addition to that two new standard initiatives were adopted recently. One is the Open Optics Camera Interface (OOCI) standard to control the camera optics through GenCP and SFNC. The second standard initiative is called emVision and aims to expand the application of GenICam into the embedded vision world combining the particularities of the machine vision industry and the embedded world.

Standards Chair discussion round

EMVA and Lakesight Technologies took the opportunity to invite all standard Chairs and Co-Chairs present in Stresa to a round table discussing how standardization has influenced the machine vision industry and what challenges lie ahead.

All Chairs agreed that GenICam has become the backbone within the standardization in the last years since nowadays all other machine vision hardware interface standards refer to GenICam as the widely established generic programming interface for all kinds of devices, framegrabber and applications.

It was also common sense that standardization has been and still is a big enabler for machine vision technology, since machine vision standards simplify adoption of machine vision technology. Standards reduce learning costs for customers and development costs for companies. This allows the vendors, the integrators, and the end users to focus on their unique aspects, and not waste resources on the mundane.

One point made in the Chair session how standardization has changed within the years was in the trend towards software standardization, whereas in former years hardware used to be in the center of standardization activities.

Looking into the future of machine vision standardization the Chairs stated that standards can never be an end in itself but instead need to undergo constant evolution both to test and ensure compatibility of new components as well as to adapt components to new and different market needs. In contrary to the successful plug-and-play approach of existing GeniCam standardization new players such as from the embedded market might be vertically layered instead of horizontally and have less need to integrate interfaces from existing component producers. Also, the current market consolidation might lead to a smaller product variety and thus to less demand for standardization.

However, cooperation amongst the existing machine vision standards but also new alliances with standards from related industries was seen to become more and more important by the Chairs. As it was put by one participant of the Chairs session: “We have to reinvent standardization again and again”; and referring to cooperation with standards from other industries: “We should be thrilled when another standard comes along to be used for a higher growing marketplace.”

Next IVSM

The International Vision Standard Meeting Spring 2020 will take place May 25 – 29 in Montreal/Canada.

About EMVA:

Founded in May 2003 in Barcelona, the European Machine Vision Association currently has about 120+ members representing more than 20 nations. Its aim is to promote the development and use of machine vision technology and to support the interests of its members – machine vision companies, research institutions and national machine vision associations. The main fields of work of EMVA are: standardization, statistics, the annual EMVA Business Conference and other networking events, European research funding, public relations and marketing. To find out more visit the web site www.emva.org.

About Lakesight Technologies:

Lakesight Technologies Holding GmbH, based in Munich, part of the Dutch stock listed company TKH Group, enables customers to fully automate processes. This is achieved by vision systems for inspection, measurement, verification, recognition, and process control via analyses of color, 2D-and 3D-shape as well as the material structure. With its six highly complementary member companies Allied Vision, Chromasens, Mikrotron, NET, SVS-Vistek and Tattile, Lakesight follows its vision to evolve as the technology leader in the machine vision industry for smart solutions. For more information visit www.lakesighttechnologies.com.

Embedded Vision Europe 2019 (eVe) Conference presents final Program

Stuttgart becomes from 24 – 25 October again hotspot for news and developments in this disruptive technology

Barcelona/Stuttgart, 16 September, 2019. From 24 to 25 October, the ICS International Congress Center Stuttgart will once again be the European focus of the embedded vision scene. For two days, the Embedded Vision Europe 2019 (eVe) conference will focus on all aspects of this disruptive technology. The conference attendees can expect an extensive range of top-notch lectures, a well-chosen selection of embedded vision solutions in the accompanying table-top exhibition and ample of time for selected networking activities.

Opening Keynote from Intel

A highlight of the conference will certainly be the opening keynote by David Austin, Sr Principal Engineer at Intel Corporation and responsible for AI-based solutions for the industrial IoT market on the topic ‘Flexible and Practical AI for Industrial Deployment’. In his speech David will present concrete steps for practical implementation options to optimize key performance metrics such as accuracy, latency and cost in the industrial use of Artificial Intelligence. He aims to enable the audience to transform the acquired knowledge directly into practice in their own IIoT projects.

The lecture by Jagan Ayyaswami from Micron Technology deals with processor architectures for machine learning. The following presentation by Ratislav Struharik from IDS titled “Universal CNN Accelerator intended for edge-base AI inference” is part of the same topic. Neil Trevett, Vice President Developer Ecosystems at NVIDIA and President of the Khronos Consortium, will give a speech on “APIs for Accelerating Vision and Inferencing: an Industry Overview of Options and Trade-offs”. Andrea Dunbar, Head of Embedded Vision Systems at CSEM joins the speaker list on the first conference day with her talk on ‘Autonomous data-logger with ULP imager’; as well as Michael Engel. The founder and president of Vision Components introduces ‘MIPI Cameras: New Standard for Embedded Vision’.

Second conference day with Deep Learning as one of the main topics

At the beginning of the second day of the conference Jonathan Hou, Chief Technology Officer at Pleora, in his lecture ‘Embedded Learning and the Evolution of Machine Vision’ looks at the rapid development of image processing in recent years. “Using Sparse Modeling in Visual Inspection to Solve Issues Deep Learning Can’t” is the topic of the talk given by Takashi Someda, Chief Technology Officer at Hacarus. Dr. Vassilis Tsagaris, CEO of IRIDA Labs, will look beyond the existing deep learning models and talk about ‘A holistic embedded vision approach: looking beyond the deep learning models’. Pierre Gutierrez, Lead machine learning researcher at Scortex talks about ‘The challenges of deploying Deep Learning for visual quality inspection’. Gion-Pitschen Gross, Product Manager at Allied Vision, addresses the practical implementation topics in his presentation “How to set up an embedded system for industrial embedded vision – Requirements, components, and solutions”. The final presentation on the second conference day will be given by Bram Senave, Business Development Manager at Easics, on the user-oriented topic of “Embedded deep learning in PCB inspection”.

The lecture program of the eVe Conference 2019 will be framed by a table-top exhibition as well as individually bookable B2B meetings during the conference breaks. Information on registration is available at www.embedded-vision-emva.org.

About EMVA:

Founded in May 2003 in Barcelona, the European Machine Vision Association currently has about 120+ members representing more than 20 nations. Its aim is to promote the development and use of machine vision technology and to support the interests of its members – machine vision companies, research institutions and national machine vision associations. The main fields of work of EMVA are: standardization, statistics, the annual EMVA Business Conference and other networking events, European research funding, public relations and marketing. To find out more visit the web site www.emva.org.