The collaboration allows the BrainMatter AI image processing system to perform automated tasks on edge processing devices in industrial environments as part of the Internet of Things (IoT). The technology can be used in a range of instances such as for the automatic inspection of products and objects, the anonymization of video images for pattern recognition and the observation of materials, terrains and buildings.
This collaboration with Amsterdam-based Braincreators enables software packages that are ready to be used in the cloud, on-premise and at the edge using widely available AI frameworks and Nvidia GPU-based edge hardware. The packages can be developed by the BrainMatter platform, which allows domain experts to transfer their knowledge to the platform based on examples. Through this transfer of knowledge BrainMatter can learn skills and perform reality checks against a given standard or objective. The actual situation, observed by the platform, can therefore be compared to a description of the situation as it should be.
The Advantech IoT devices execute BrainMatter's automation models in near real-time as the data being processed closer to the device, reducing network latency. In addition, the data is processed locally, reducing the amount of sensitive data traffic and ensuring customer security while saving energy.
"Our partnership with BrainCreators represents an important strategic partnership for Advantech. The partnership allows us to better roll out our AI solutions in the Netherlands and beyond," said Jash Bansidhar, Managing Director of Advantech Europe. "As Edge AI becomes increasingly important, it is vital that we work with partners who are able to combine domain knowledge with a practical mindset to deliver successful projects. We are convinced that BrainCreators can help customers get the best out of Advantech's Nvidia technologies.”
“Advantech has an enormous amount of knowledge and technology in the field of edge computing. By combining their edge technology and knowledge with BrainMatter, we have developed together an end-to-end solution that enables visual inspection and observation in almost real-time," said