New software enables real-time analysis of the interior of complex products using 3D image reconstruction to detect deviations or production errors, without damaging the object. This makes it possible, among other things, to detect rotten apples or errors in the assembly of electro-mechanical products as they pass by on the assembly line.
Quality inspections typically involve random checking of products, checking a few products to check the quality of a large number of products. In practice, this means that by no means every product is inspected. If production problems are discovered at a later stage, it is often necessary to recall a large number of products, which are taken out of circulation and in many cases can no longer be sold. iXCon offers a solution by making it possible to detect production errors in a time- and cost-efficient way, without damaging products in the process. This makes it possible to check many more or even all products in a production run.
iXCon
The iXCon project was carried out by imec in collaboration with UAntwerpen, KULeuven, UGent, Robovision, Terumo and XRE. The project was led by VisionLab, an imec research group affiliated with the UAntwerpen. Within the iXCon project, research is conducted on behalf of and in close cooperation with Flemish industrial partners.
iXCon has led to the development of three concrete applications. These include a 3D X-ray inspection that allows inspection of one product per second. In addition, two agricultural applications have been developed for quality inspection of harvested apples and automated crop planting, respectively. The applications were developed based on the ASTRA toolbox, a 3D tomography framework developed by VisionLab.
Applications in practice
The following applications have already been deployed in practice:
- Inspecting flower bulbs: Technology developed within iXCon has been deployed to check the internal structure and quality of flower bulbs. This has increased the efficiency with which flower bulbs are planned automatically by ensuring not only that all planted flower bulbs are of sufficient quality, but also that all bulbs are planted with the right side up.
- Non-destructive internal inspection of fresh fruit: Using technology developed within the iXCon project, pears have been sorted in an automated way after harvest. This helps create more homogeneous loads and also reduces waste.
- Quality control of small medical devices: Technology developed within the iXCon project has been deployed for quality inspections of high-volume medical devices. The technology should eventually be able to be integrated with existing inspection techniques that manufacturers have been deploying for some time.
X-ray
Current applications are developed for inspection using X-rays. However, the software can also be used in combination with other imaging methods, such as visual, infrared or hyperspectral inspection. Users can specify information about the product to be scanned in advance to increase the quality and speed of inspections. Technical drawings can be used, for example, and it is possible to enter data on the shape and composition of apples.
Henk Houf, COO/VP Manufacturing & Operations Terumo EMEA: "Being able to deploy this new innovative technology will definitely allow us to take our quality inspections to an even higher level. An important aspect for Terumo with this technology is its flexibility and deployability in an industrial, high speed, high quality environment and this for several of our products."
Still under development
The software is still under further development. For instance, the parties involved want to build a hardware-based system that offers the flexibility to scan several different products in a short time. In addition, the hardware setup is expected to be able to be used to develop and validate new measurement methods.
Author: Wouter Hoeffnagel
Source: imec (pictured)
