With the Hawkeye project, Brantner has redefined the standards in the image-based recognition of objects in the waste management sector. With the support of our neural network, we are currently able to distinguish among 32 different contaminants in real time in the bio-truck's bulk and assign them a school grade from 1-5.
Together with our biologists from the Business Development Team, a "hazard matrix" was created, which is used as the basis for evaluating each individual ton. Not every interfering substance is equally "harmful" or "dangerous". A single PET bottle is less "disturbing" than an entire car battery.
From all the individual pours, a tour score per truck can be made. Based on this average total score, we can dynamically route our press trucks to the right storage location. There, the load is sorted again as required or goes directly into the production chain of our premium composting plant "Erdenreich".
When implementing our innovative projects, we do not only focus on technology, but also on its usability. Our system works 100% autonomously which makes futher interaction from our employees totally unecessary. Thus, 'Hawkeye' does not prolong our tours, rather, it makes them even more efficient.
Thanks to our "rapid prototyping" process, we were able to produce the necessary housings ourselves quickly and cost-effectively using our own 3D printer made from recycled filaments, and react efficiently to any necessary adjustments.
We are currently collecting more and more photos from the trucks and forward them on to our neural network, which is getting better and better every week by means of machine learning.
Our experienced AI trainers take care of the quality assurance of the neural network.
