Automating and scaling visual inspection with AI & Machine Learning

Our software is automating and scaling visual inspection and defect detection on metals and plastics. We are leveraging the state of the art in computer vision and machine learning.

Get in touch

Our software is automating and scaling visual inspection and defect detection on metals and plastics. We are leveraging the state of the art in computer vision and machine learning.

Trusted by customers

GKN uses our technology to get more accurate classification of defects and to scale visual inspection quickly from one part to another and from one production line to many.

Alupress uses our technology to speed-up cycle times on complex parts by leveraging our 3D based machine learning approach. Furthermore, Alupress increases classification accuracy and saves costs.

Aluflexpack - a leading international manufacturer of flexible packaging - is leveraging our machine learning based visual inspection software to automate end-of-line quality control. The collaboration started in the plants in Croatia.

Toyota is one of the largest global car manufacturers. We are working with Toyota Motor Europe to leverage our unsupervised machine learning software to automate visual inspection on metal parts.

Swarovski is a globally recognized producer of crystal glass and jewelry. We work with the company to leverage our unsupervised machine learning approach to have Swarovski employees train new models quickly and hence scale reliably to many production lines.

Hydac is an international company group that specializes in production and distribution of components and systems related to hydraulics and fluidics. We work with Hydac on analyzing the usage of supervised and unsupervised ML approaches.

Our solution


No human required for lengthy, week-long programming of visual inspection systems. We leverage our unsupervised learning approach.


Higher detection rate compared to inspection by humans and traditional vision systems. Pseudo-scrap rate is reduced.


Defect detection learnings are transferred from one to many products on many production lines.


Free 30min consultation video call

Helps us understand your quality control and defect detection use case and gives you an idea of our solution and offering.

Image capturing

We work with images from your production line. We integrate with your current vision system or can help you build your own specialized vision system.

Feasibility study

To start we need 5.000 images of good products with no defects and 500 images of products with defects to establish proof-of-concept within 2 months.


If our software proves to work, we can deploy it inline, expand the solution to more products and more production lines leveraging a flexible Software-as-a-Service model. Only pay for what you need.

Areas of application

Our technology can be used to automate the end-of-line quality control process in various downstream industries that leverage non-ferrous metals such as aluminium, copper, and zinc as well as ferrous materials such as steel: construction, automotive, mechanical and electrical engineering, aerospace, medical devices.

Independent of the industry where our technology is applied, factory operators are using our software on a workstation that is installed next to production lines. The interface allows them to check the performance of various production lines and parts simultaneously, to start and stop production sessions and to report parts and results to managers.

Managers on their end are able to get a holistic overview of production and quality statistics in real time, across production lines, facilities and subsidiaries.

Management Panel

The Management panel groups and displays all relevant information in one, easy to access online web-app. Decision makers like quality control managers, operators, CXO’s can easily log on and check quality of parts being produced in real time: for multiple live parts being produced on different production lines for different customers in different plants. The management panel is synced to the in-line system and can therefore be used to also look at the parts, defects and issues that were highlighted and commented by the operators at the shopfloor.


Covision Quality is a spin-off of Covision Lab and works with its research and institutional partners. Furthermore, Covision Quality is part of Silicon Valley and Munich based Alchemist Accelerator, the world’s prime enterprise accelerator, and NVIDIA’s inception program, the leading hub for A.I. startups.

Download our white paper

Download our whitepaper if you want to know more about Covision Quality’s approach to automating and scaling visual inspection.

Please fill in this mandatory field.
Please fill in this mandatory field.
Please fill in this mandatory field.
Please fill in this mandatory field.
Please fill in this mandatory field.
Please agree to our privacy policy.
Your request has been sent with success.
Unfortunately we could not send your request successfully. Please try again later.


A team of experienced and visionary technology entrepreneurs works with best-in-class computer vision, artificial intelligence and machine learning researchers and engineers to disrupt the visual inspection market – starting with metals and plastics. We are currently looking for Computer Vision Engineers and Strategy and Business Development Interns.

Join our team by sending your resume to

Trade Shows & Events

Featured In