Innovation is the key to staying ahead of the rest. Our customers know this and are constantly searching for new ways to improve their efficiency or the quality of their products. We are there to take innovation on step further. So if your hardware vendor or IT partner tell you that something is impossible, please talk to us.
Combining Artificial Intelligence with computer vision is like giving a camera brains of its own. It would be able to learn by example and it would be able to deal with ill-defined objects or scenes, making it especially suitable when dealing with organic products, people or very noise data.
Using a unique proprietary algorithm, we are able to detect objects based on their CAD drawing. The algorithm calculates the position and rotation of one or more objects, which can be used at various points in production lines. Switching to a new product can be done instantly, without the need to train the algorithm or physically adjust your hardware.
Adding a camera to your production line will allow you automatically spot defects in texture, size, positioning or color. It is even possible to mount one or more cameras on a robotic arm to achieve 360 degree inspection.
CCTV is currently used extensively on public and industrial sites, but as the number of cameras increases, it is becoming impossible for human operators to monitor all streams and give them their full attention. Through automatic analysis of the camera streams, it is possible to trigger alarms or sort the streams by relevance, helping the human operators to respond faster and miss fewer key events.
Identifying and sorting products using serial numbers or barcodes might seem like a talk that does not require Artificial Intelligence or other complex algoritms. However, the typical off-the-shelf products available for OCR and barcode scanning will not work well if conditions aren't optimal. We have developed technology to succesfully to run text recognition and barcode scanning on mobile devices, on highly reflective objects and on curved object.
Not only do we develop provide smart vision software - the brains of your automation solution, we also provide a number of complementary services which allows us to provide you with a complete custom hardware/software package.
We can provide complete automation solutions for pick-and-place, product inspection, sorting and other automation processes. We can do this in collaboration with our robotics and automation partner, or working together with trusted partners of your own.
To support our vision applications, we regularly build custom lighting solutions with better specs and at a lower cost than off-the-shelf solutions.
We have helped helped several customers design, program and test custom PCB boards. This included wireless communication, sensor neasurment logging and high-efficiency LED driver circuits. We have the tools and expertise needed to develop software for a number of MCUs, including ATTiny and ATMega chips.
We are skilled in both ASP.net and PHP web development and have partnered with a number of freelancers that can further assist. This allows us to provide web-based solutions around our innovative technology.
We are a small and efficient organization, structured as a network of dedicated freelance professionals.
Hans van Kuilenburg, founder and CEO of VanAI has a master in Cognitive Artificial Intelligence and has over a decade of experienced in the application of computer vision and general AI techniques.
Several freelancers working through VanAI have extensive AI, mathematical or computer vision knowledge, but we also have professional web developers, designers, text-writers, database engineers, mobile phone developers and mechanical engineers in our network.
We are located in Scheveningen, The Hague. Come visit us for a cup of coffee in our historical office building or on the famous Scheveningen boulevard.
We are proficient in using Neural Networks, Support Vector Machines and other supervised learning algorithms for classification tasks. (E.g. labeling an inspected product as good or damaged)
A 3D depth image an be obtained through a special 3D camera, but also by combining the camera image of two regular cameras or by projecting a laser pattern. The resulting depth information can be critical in various applications, like bin-picking or sorting objects by shape.
Over the years, we have developed several proprietary algorithms and methodologies. E.g. we developed a way to match objects using only a CAD drawing and have a tool-chain for reliably classifying object surface textures.
We prefer Microsoft .NET as our development platform, which gives us excellent debug capabilities through Visual Studio, a rich and stable .Net base library and rapid development without compromising on speed by using a combination of C# and C++ as development languages.
Halcon is the leading commercial library for Vision applications. It enables the rapid development of reliable applications. We are trained and experienced in the use of Halcon.
Various open source vision libraries exist which can be used in commercial applications. OpenCV is such a library and has been improved over many years by a large group of passionate developers. Especially in applications where it is essential to have full access to all source code, using such a library can be benefitial.