Fotios Konstantinidis

“I-SENSE is one of the Research Groups of the Institute of Communication and Computer Systems (ICCS) of the National University of Athens in Greece (NTUA). Founded in 2002, I-SENSE Research Group is committed to the highest standards of academic excellence and research innovation in the field of electrical, electronic and computer engineering & technologies. For the process of sorting and categorizing materials, I-SENSE group provides an innovative robotic system that integrates multiple sensing modules and advanced machine learning and deep learning algorithms to accurately separate different types of materials/wastes. This type of system can upgrade the outcome of waste streams as it separates materials based on their composition, size, or any other desirable property. More specifically, this sorting system consists of a conveyor belt and a pre-treatment mechanical screening system. The screening system ensures a uniform distribution of the extracted materials on the conveyor belt that conveys the materials through a multi-sensory module. This module consists of RGB cameras that provide information regarding the size, shape, and position of the material on the conveyor belt. Hyperspectral cameras can also be embedded to provide information about properties of the materials that do not arise from the visible spectrum. X-ray screening sensors may also be used for additional information as well as characterization techniques that give information about the chemical composition of the material, such as X-ray fluorescence (XRF). Machine and deep learning models can analyze all the extracted data gathered from the sensors, separately or in combination, and classify the corresponding material. After classification, robotic modules or air nozzles (depending on the size of items) separate the materials based on the required properties. The operational activities are supported by an Extended Reality application through a Head Mounted Device. The XR application aims to enhance the efficiency of waste sorting activities by providing real-time visual cues and information overlays to guide the human operator, improving accuracy in sorting activities. Finally, our demonstration will include an additional XR application coupled with ML algorithms, targeting novice practitioners (such as consumers and unskilled workers) towards intuitive sorting and recycling _We provide adaptable industrial sorters for:   a) Plastic Sorting (Fossil-based, Bio-based) b) Wood Sorting c) Metal Sorting d) Textile Sorting e) Mines/Ore characterization _