Artificial intelligence robot sorting technology is the future of a centralized, data-driven recycling model. Usually combined with optical sorting technology to get the best results, the latest robotic sorters are achieving (and starting to exceed) 70 sorts of mixed plastic and paper per minute. They are also very effective for C&D classification, and they have great potential in other materials, including metals, electronics, organics, and glass. On the one hand, due to rapidly changing and highly polluting feed streams, as well as the shrinking of global end markets such as China, and the underdeveloped domestic market for many recycled materials, the pollution of final products has become the primary problem for MRF operators. The level of purity required by today’s end markets is one of the most significant challenges facing recyclers. The use of artificial intelligence (AI) robotic sorting technology has increased production and recycling rates, and improved the quality of recycled materials—improving its ability to compete with original products on the market. It is also important that manual pickers cannot achieve close to 70 picks per hour on average, and they may be unreliable, sick or injured. We also know that in the very difficult working environment of recycling facilities, the staff turnover rate is very high. Robot sorting helps to solve these problems, because the robot can work as much as needed and pick reliably, without being affected by dust, hazardous materials and other risk factors in the recycling plant.
Machinex’s SamurAI, launched in 2018, is a four-joint robot that uses artificial intelligence technology to identify materials for accurate and active product recycling, or as an accurate quality control function. The artificial intelligence of the system runs in a pre-determined order of task levels to maximize financial returns, while continuously improving and learning operating experience to ensure maximum recognition efficiency. According to Machinex, SamurAI can double this average with up to 70 picks per minute compared to a manual sorter with an average of 35 picks per minute. SamurAI is also designed to accommodate sorting conveyors up to 48 inches wide, while providing a modular design for multiple robot configurations. Machinex recently launched two new high-tech products-MACH Vision and Intell platform-both are compatible with SamurAI integrated robots and MACH Hyspec color sorters and other high-tech systems for MRF. MACH Vision is a data acquisition station driven by artificial intelligence. It can combine field-proven technologies such as infrared hyperspectral technology, visual recognition and metal detectors to analyze fast-moving material flows very accurately. The new Intell platform is directly connected to all the above-mentioned high-tech automated sorting equipment, providing operators with real-time material flow data capture and analysis functions.
Waste Robots developed their WAR (Waste Autonomous Recycling) software to allow multi-sensor scanning and real-time artificial intelligence analysis, including waste chemical composition and shape recognition. WAR is used with the robot technology manufactured by FANUC to calculate the grasping strategy and picking sequence in real time to execute the robot to extract recyclable materials and provide the ability to disassemble the debris accumulated in a line. The WAR system combines 2D and 3D color and hyperspectral object chemical recognition with a vision system that adapts to the width of the conveyor, and it allows multiple robots to be placed on the same sorting line (each additional robot increases by about 20% according to the waste According to robotics companies, it improves the efficiency of a given production line). The technology combines computer vision and deep learning algorithms with state-of-the-art robotics technology to achieve smaller, more accurate, safer and more profitable classifications in organic recycling, MRF applications, and C&D recycling.
Bollegraaf’s AI-driven RobB-AQC is a fully automatic robotic sorter designed to improve sorting quality control. According to Van Dyk Recycling Solutions, Bollegraaf’s North American distributor and product support company, this is the first sorting robot that combines the accuracy of NIR detection with the adaptability of artificial intelligence learning and serves as the final quality control step for the container line , One RobB device can recycle up to 70 picks per minute-achieving higher productivity and better reliability than two manual sorters. RoBB also uses laser guidance systems and cameras to detect the height of waste products, combined with near-infrared, to achieve accurate classification of recyclables through material recognition and 3D detection. ROBB-AQC is designed with flexibility in mind, and can be placed on top of the existing sorting line with extremely low transformation costs and shortest downtime. Material selection is just a button, ranging from PET, HDPE, LDPE, PS and PP to Tetra Pak, OCC or paper/cardboard of various shapes and sizes.
In 2021, AMP Robotics launched an AI-driven material characterization software solution called AMP Clarity, which can identify and classify recyclables that flow through different stages of the recycling process. The company stated that this latest industry innovation is a breakthrough that aims to provide measurable transparency for the recyclables captured and missed in different recycling processes, and to confirm the value for resale to the end market of the supply chain. Recycle the components of the material package. AMP Clarity collects data on mixed plastics, including PET, HDPE, LDPE, PP and PS; aluminum beverage cans (UBC); and fibers such as corrugated cardboard (OCC) and classified residential paper and newsprint (SRPN). Material data is collected, sorted and subdivided so that MRF can see which recyclables are flowing at different stages of its operations. According to AMP, this monitoring helps prevent recyclables from being lost to landfills, diagnose gaps in processing efficiency, review the composition of the material flow, and determine the final bundle content and material quality for buyers and sellers.
FANUC is one of the largest robot manufacturers in the world. Its business covers multiple fields. It has recently started to establish a partnership with Recycleye, an artificial intelligence expert start-up company in the United Kingdom and France. In the past year, Recycleye has combined its proven artificial intelligence computer vision technology with FANUC’s experience in robotic manufacturing and automation to deploy and expand Recycleye Robotics’ artificial intelligence-driven robotic waste collection system. FANUC engineers designed Recycleye Robotics to be 75% lighter than any existing robotic garbage collector on the market, allowing plug-and-play installation, thereby eliminating traditional expensive retrofit costs. According to Recycleye, the system performs the physical tasks of identifying, picking and placing materials at a rate of 55 times per minute, automating traditional manual operations and doubling the overall throughput of the facility.
In 2019, Bulk Handling Systems (BHS) launched Max-AI AQC-C, the solution consists of Max-AI VIS (Visual Recognition System) and at least one collaborative robot (CoBot). CoBots are designed to work safely with people, which allows AQC-C to be quickly and easily placed into existing MRFs. According to BHS, Max-AI technology recognizes recyclables in a human-like manner, guides robots and optical sorters, and collects and reports material characteristics. The Max-AI AQC (for autonomous quality control) robotic sorter is ideally suited to be placed at the last opportunity position on the production line to identify and capture the remaining recyclables before they exit the process, and/or with the NRT SpydIR optical sorter The combined use effect is very good to identify and remove the non-fiber in the upstream mixed paper products. In addition, the latest AQC-C solution can be installed in sorting cabins, narrow aisles and other cramped locations, and is easy to expand-allowing up to four robotic sorters to be added behind each Max-VIS system.
Steinert’s new UniSort Unibot combines reliable sensor technology with modern robotics and artificial intelligence-based recognition software to ensure accurate detection and pure plastic parts in the final sorting step. UniSort Unibot relies on a sensor combination composed of high-resolution NIR and color cameras, and uses optimized detection software for fast data processing, even under high belt loads. This technology relies on the use of intelligent object identifiers based on artificial intelligence, and uses delta robots with specially developed tool adapters to provide the fastest speed and maximum accuracy in a very small space. According to Steinert, UniSort Unibot checks and simultaneously evaluates the quality of the material flow, which guarantees the best quality of the final product. It also creates an interface for digital processing of data so that it can react to the material flow immediately to meet the specific requirements of regulations and customers.
According to the manufacturer, EverestLabs has released its most compact, high-recovery robotic footprint unit for MRF applications, which can pick up up to 208 times per minute, a 49% increase in object picking compared to traditional Delta robotic solutions . EverestLabs stated that their robots have successfully achieved a picking success rate of 85% or higher due to their innovative arm-end tools, software and suction power-which is by far the highest production robot in MRF. These robotic cells are also customizable, highly flexible, and use automation, operational intelligence, and ERP/ESG components powered by proprietary RecycleOS technology. These robotic cells also use patent-pending software with low inference time and can adapt to various MRF belt configurations, including slope, low overhead clearance, trough and V-shaped MRF belts.
The ZenRobotics fast picker was launched in 2018. It is the culmination of ten years of experience in the field of garbage robots and combines artificial intelligence software with high-speed picking. Fast Picker is designed to increase the efficiency of MRF and is ideal for lightweight materials such as packaging (LWP) and dry mix recyclables (DMR). According to ZenRobotics, the technology can perform up to 6,000 picks in a three-arm configuration per hour. ZenRobotics’ fast pickers allow 24/7 automatic picking and increase recovery rates while maintaining high purity of recyclables. It is easy to integrate into side stream, waste recycling line and quality control after optical sorting. It has a sturdy and compact design that can be retrofitted to different conveyor widths, and most picking stations can be installed without additional modification. ZenRobotics Fast Picker is supported by ZenbrAIn, a unique artificial intelligence software that analyzes data and controls robots.
This article was originally published in the July/August 2021 Recycling Products News, Volume 29, Issue 5.
Post time: Sep-10-2021