A new photonic standard component for3D sensing

Informations projet

K-IND

N° de convention de subvention: 954933

DOI

10.3030/954933

Projet clôturé

Date de signature de la CE 10 Juin 2020

Date de début 1 Juillet 2020

Date de fin 31 Janvier 2023

Financé au titre de

• INDUSTRIAL LEADERSHIP - Innovation In SMEs

Coût total

€ 2 229 375,00

Contribution de l’UE

€ 1 560 562,50

Coordonné par

K LENS GMBH
Germany

Periodic Reporting for period 2 - K-IND (A new photonic standard component for3D sensing)

Période du rapport: 2021-07-01 au 2023-01-31

Résumé du contexte et des objectifs généraux du projet

Automation of industrial production promises cost savings, quality assurance and the avoidance of inhumane working conditions in industrial production plants. But despite technological developments over the past 40 years, many industrial tasks remain, where the automation industry is still struggling. Modern sensor technologies, in particular photonic machine vision sensors, offer solution potential. And while traditional vision sensors only provide 2D information, several approaches have been developed to record 3D. But even with state of the art 3D sensors, the human eye cannot be replaced in many cases due to the complexity of products, materials, lighting conditions and shortcomings of existing technology making automation impossible.

Our patented light-field lens technology addresses such flaws. It is small, video-capable, compatible with existing technology, robust even in complex lighting conditions and precise. The hardware is a new optical component (the K|Lens) which can be used with any standard camera and any main lens turning the system automatically into a 3D system.

The software is based on latest research results in the areas of computer vision and machine learning.

Travail effectué depuis le début du projet jusqu’à la fin de la période considérée dans le rapport et principaux résultats atteints jusqu’à présent

As a result of the project we want to develop within 2 years a first portfolio of solutions comprising both hardware (optical components) and the related software and prove the value of the technology in first POC customer projects.

After year 1, the hardware is mostly specified and first versions are transferred into production. The software has been developed to the extent that it can be tested with customers in the field. It comprises both classical mathematical approaches as well as modern machine learning concepts (CNNs).

The development is accompanied by continuous evaluation tests with interested partners. Overall, more than 20 feasibility studies, mostly with manufacturing companies have been performed showing generally positive results. This market validation is now extended towards system integrators and camera makers to leverage the commercialization strategy.

The use cases "bin picking" and quality control of small parts, for example in the electronics industry, still show the highest potential.

A particularly strong market demand can be perceived by our well balanced 3D solution between cost and quality.

Progrès au-delà de l’état des connaissances et impact potentiel prévu (y compris l’impact socio-économique et les conséquences sociétales plus larges du projet jusqu’à présent)

What we have been able to show is that with a multi view approach, in particular complex surfaces like glass or (semi)-transparent plastic packaging can be properly detected by 3D algorithms. Similarly, for small and complex structures a complete 3D visualization of the interior life has been shown which has not been available before.
Until the end of the project, the hardware software solution will be tested intensively in the field, so that market validation is completely obtained.

K Lens test on conveyor for electronic parts