Success is no accident, it's hard work, perseverance, learning, studying, sacrifice and most of all, love of what you are doing or learning. Without that passion the Robovision team would not be where it is now.- Philippe Muyters, Flemish Minister for Work, Economy and Innovation
Robovision changed the diamond cutting process worldwide. Even in China companies are willing to buy this excellent engine. So far we have sold 180 units worldwide. The software is an important reason to buy our hardware...- Eddy Giron, ADT
The line and integration is running very smoothly. The girls working the line no longer have to label the trays and no longer have to scan them into Picas to verify that they were started. We are also tracking the performance of the robots per genus, so we have some stats to know how many they should expect to be done per hour.- John Stallmer, PICAS


Learning arbitrary plants, the AI way

We use deep neural networks to teach a robot arbitrary new plants. All you have to do, is put a few samples on the conveyer belt, use our RVAI portal to create a model, and in the next few minutes the robot is ready for planting a new plant type. Next time you buy flowers for your loved one, think about our deep learning engine.

The video shows an installation in Carleton, Michigan.

Deep Learning in the car industry

A solid and validated flow of components is essential for a lean car assembly process. High tech scanning technology can guarantee that a specific component will arrive just in time on the production line. Audi relies on Robovision for this crucial part of their process..

By applying 3D deep learning to manufactured pieces, Robovision checks whether the right subtype is selected, and whether a set of components is complete for a specific car type. In this case RVAI ensures that the WZB (the kit you use when you have a puncture) is complete and validated. Thanks to the strong GPUs from NVIDIA we keep the cycle time fast and snappy.

Deep learning in the packaging industry

Pepsico manufactures and packages Dorito chips in the highlands north of Mexico City. Package branding is changing so frequently that a powerful AI based central system is needed to teach in new branding very fast. Robovision‘s RVAI is the solution here. Reflective, blown up and unpredictable, recognising branding is not for the faint hearted in this application.

AI applied to 3D in Agriculture

The next step in our agricultural developments is the art of convolutional networks applied to 3D.  Now every day a Jumbo 747  flies in from Africa full of Chrysanthemum stems. We want to grow them here in Western Europe instead and manipulate them with smart robots (that pick the stems from the mother plant). This way the fertile ground of Ethiopia can be used to feed the hungry in the region. The AI revolution can generate powerful secondary effects, disrupting the flow to cheap labour and create a more harmonious world order.

AI applied to the surface inspection done by scanning electron microscopy

We apply machine learning algorithms to inspect surfaces that underwent a special nanotechnological surface treatment. The exact degree of corrosivity is essential in the subsequent process, and the quality of the final product highly depends on it. The devil is in the details.

AI applied to thermal streams

Chemelot , one of the biggest chemical sites in Europe, relies  on Robovision to detect ammonia leaks . Our Gashawk™ software uses state of the art algorithms to distinguish between endothermal reactions related to ammonia, and other similar phenomena.

For more compact projects, we use embedded technology to run AI on Flir cameras. With techniques such as squeezenet and noscope, we optimise complex architectures for low power devices. ThermalHawk™ is the platform we offer here.

Multimodal deep learning applied to xray 

One of the main advantages of deep learning lies in the amount of complex data that can be aggregated and used as input for the deep architecture. Combining hyperspectral imaging, 3D point clouds and X-ray can generate tons of  information, allowing for interesting chunks of data to be combined in an intermodal way (see more in this video).

Together with IMEC and Fraunhofer we invest in the future of deep learning, tuning deep architectures to rapidly process time variant 3D voxel data.

Deep learning based avatar search in urban environments

(this video is intentionally blurred because of NDA and privacy reasons )

We have connected our deep learning engine to the Genetec platform, and now we can analyse security streams in realtime to look for people with specific traits or clothing. Imagine a missing child in an urban environment. The clock is ticking, instant action is crucial. Thanks to a powerful IBM backbone we can perform a search in realtime.

Automated pharmacies

Our software is already being used to run the fully automated pharmacies of the future. We drive the image processing behind the scenes, which enables automated recognition and processing of  medical bottles, boxes or whatever type of container.

Automated creation of UDP

Together with R.I.S. we have developed machinery for unit-dose packaging. A rig of cameras automatically detects pills of all kinds and the robot glues these on to a strip. This has proven to be a very useful aid in high tech hospitals.

Automated warehousing

Vision and image processing allows for fully automated warehousing. Here our systems are critical for safety, recognising the type of stones and the size of the pallet before it goes into the ‘Manhattan’ warehouse. Xella relies on Robovision for this kind of critical, image processing based use cases.

Image processing in the food industry

Currently we are combining hyperspectral imaging and deep learning in the food processing industry, these tracks are still in progress (see the explanatory images below about the principles behind hyperspectral imaging)

Regarding the video at the bottom of the page: this is an older, discontinued project (2011) where more usual image processing is applied (image segmentation). Nevertheless we show it here to demonstrate our ability to efficiently build prototypes for different industry niches.

Many more case stories:

  • We are using deep architectures to get to the bottom of hyperspectral imaging.
  • We are using a special brand of convolutional neural networks (LSTM networks) to detect anomalies in cyber security related logfiles (for now with limited success, still a long way ahead, images and video have no secrets for us, text files however…).
  • We are looking into FPGA and embedded development to keep our edge and to remain in the blue ocean part of this rapidly emerging new business.
  • We applied eulerian video magnification to read heartbeats out of normal video streams.