国产麻豆

Automating the process, optical sorters boost throughput, recovery, and purity if the right investment is made.
By Sebastian Ward

Optical sorters are designed to automate the process while increasing yield, purity, and profitability. Increasing throughput, optical sorters using valve block ejection deliver a much higher capacity than that of manual labor, helping to address the labor shortage and allowing for reallocation of workers from the line to positions of more value for the materials recovery facility (MRF).

However, investing in the right technologies, partnering with an experienced provider, and maintaining the equipment are all keys to improving ROI and the operation鈥檚 profitability. It all begins with having clearly defined goals and investment targets, must-have and nice-to-have requirements, as well as budget framework.

Working with a trusted partner to help guide you through the necessary considerations鈥攆rom sensor and technology selection, feed material testing, and legislation and market requirements鈥攊s the best place to start. They can make recommendations for flexible plant configurations that include advanced technologies capable of adapting to changing market conditions and legislation, helping to ensure the MRF鈥檚 profitability not only today, but also well into the future.
Following are some considerations when investing in a new circuit or making technology upgrades, as well as how to keep it operating efficiently for years.

The Right Partner
Investing in the right sorter reaches beyond finding the right sensors and technologies. It begins with choosing the right partner who will work with you to provide training, service, and maintenance after the purchase.
Today, it is critical to select a partner that can help you navigate the differences between traditional artificial intelligence (AI) used in optical sorters, and deep learning AI employed by the most advanced sorters. Both have their own unique values to the operation. Near infrared (NIR) and other sensors commonly used at MRFs sort based on material type, distinguishing between polyethylene terephthalate (PET), polyethylene (PE) and polypropylene (PP), and other plastics. Deep learning, on the other hand, sorts based on visual information, such as material color and what the eye can see.

During equipment and technology installation, it will also be the provider that will carry out inspection, recalibration, and optimization services, involving your technical team along the way. They will also be responsible for training your team machine operation, changes to settings, as well as basic maintenance. Look for a provider that will also offer the option of remote maintenance to help save time and money.

Passing the Test
Testing infeed material in advance will provide deep insights into the performance capabilities of the proposed combination of sorting technologies. Especially when working with unusual material or non-standard applications, taking the time for this step provides significant advantages in understanding the quality and quantity of the output material.

This test may be conducted in the plant builder or technology provider鈥檚 test center, via a visit to a plant processing similar material or through a mixture of both. If processing material at a test center, your partner will analyze the material composition and run different tests, handling steps and technologies to evaluate throughput and purity rates. This will allow you to zero in on the right type, number, and location of the sorters. From throughput capacity to application feasibility, you will come away with firsthand experience and evidential data to create a sound, value-based investment strategy.

Anticipating the Future
Legislative changes plus the dynamic recycling industry make long-term planning indispensable. Even if the solution meets your requirements today, clarify with your partner in advance what options are available for upgrading the sorter.
We have often received requests from customers a few years after plant installation seeking new opportunities to enhance their throughput, purity, and/or recovery to capture new revenue streams through plant retrofitting. For example, a process using NIR sensors alone was determined to be the ideal solution to process tested infeed material at the time of installation, but market and/or stream changes have the customer considering separation by color type or a solution to address high black content in the material. Having an adaptable sorter capable of add-on technologies can save considerable time and money for the MRF.

Fortunately, some optical sorters available on the market offer upgrade flexibility and a variety of sensor upgrade options. These could include adding laser technology to detect black material or, popular today, installing deep learning AI to deliver a more granular sort of final material.

Combining AI Technologies
The introduction of deep learning AI to the recycling industry has enabled the separation of previously hard-or impossible-to-classify materials. Early forms of deep learning applications included a color camera combined with robotic arms designed to mimic the picking action of the manual sorter, offering around 70 or 80 picks per minute. They were end-of-line, final quality control solutions to enhance material recovery. A drawback, however, is that the suction cups on the robotic arms can require virtually daily maintenance.

More recently, technology providers have introduced multiple deep learning AI applications to enhance the sorting capabilities of optical sorters. The advantage is twofold. First, valve-block ejection employed with sorters significantly increases throughput speed well beyond the 70 to 80 picks per minute to rates reaching 6 tons per hour as purity exceeds 95 percent.

Second, the color camera used with deep learning sees color, shape, and other visual characteristics of the material on the belt. The NIR sensor of the optical sorter distinguishes the type of material. In plastics sorting, deep learning will recognize a green soda bottle, clear water bottle, or milk jug passing through the sensor, while the optical sorter鈥檚 NIR sensor will simultaneously distinguish if that object is a PET bottle with either a PET or PVC label.

There are, however, profitable applications for the camera plus deep learning AI alone when combined with the speed of valve-block ejection. In used beverage container (UBC) classification applications, trained deep learning AI technology quickly distinguishes between the desired UBC fractions and non-UBC materials like aluminum bottles, food cans, and trays on the belt. Valve-block ejection in this application offers high-throughput processing at up to 2,000 ejections per minute and 98 percent+ purity without manual sorting.

Consider the Data
Sensor-based sorters generate vast amounts of data about the scanned material鈥攖hroughput, material and size distribution, acceptance and rejection rates, among others鈥攊n which MRFs can leverage to make fact-based decision making. It can be used to enhance operational efficiency and sorting output, while, at the same time, reduce sorting interruptions and downtime. Data enables close monitoring of input feed material, detection of unwanted material, and the ability to act quickly to correct unforeseen issues.

To gain the full potential from the sorter, work with the technology provider to connect the equipment to a secure, cloud-based platform to receive near-rear-time monitoring. This allows you to access digital metrics from a desktop computer or mobile device on the status and performance of the sorting equipment from anywhere an internet connection is available. Customizable alarms and reports also allow for close monitoring of the sorting process to quickly detect low throughput rates or purity, so it can be addressed.

Maintaining the Equipment
Today鈥檚 sorting equipment are finely tuned machines that help to automate the sorting process. To keep these machines working at peak performance, they must be properly maintained. Some of the most common maintenance practices to help keep a consistent sort include:
鈥� Calibration鈥擳o see material correctly, it is crucial that the sorter鈥檚 illumination unit is calibrated, and this sometimes needs to be done on a daily or weekly basis. Look for optical sorters that offer continuous calibration, so calibrating the machines does not cut into the production schedule.
鈥� Valve checks鈥擳o work properly, the valves in valve-block ejection units must remain clean and clear of debris. Have workers perform a valve check for high and low pressure to make sure each valve is working properly to avoid lost output or contaminated final product. Some, but not all, optical sorters offer auto testing and cleaning of the valves to ensure optimal sorting performance and purity.
鈥� Belt speed鈥擳he correct belt speed is the difference between high product purity or increased material contamination. Belt wear, bearings, gears, and debris build-up can change belt speed, so it is essential to check for correct belt speed multiple times each shift. Sensors can be used on the material feeder to assist with this.
鈥� Keep a spare鈥擳o minimize downtime, keep enough critical and high attrition parts鈥攁ir filters, lamps, valves, etc.鈥攐n hand for quick replacement. A bulb is a 10-minute replacement if in stock, but, if not, it will be one to two days to replace plus the cost of overnight shipping and circuit downtime.
鈥� Leverage technology鈥擜 connected asset enables manufacturers to resolve many machine sorting issues remotely via a virtual private network (VPN). With a call, text, or e-mail, you can connect with the manufacturer to resolve performance issues quicker than a site visit.

Maximize Your Investment
Recycling facilities today have more advanced sorting options available to them than at any other time previously to help achieve throughput, purity and recovery targets for processing infeed material. In order to maximize optical sorter investment, it is important for the MRF to have a plan with clearly defined targets, and test feed material prior to purchasing the sorters. Work with a trusted and experienced plant builder or technology provider who will not only provide you with expert installation, training, and optimization of the sorters, but will also be there to support the equipment well into the future. | WA

Sebastian Ward is TOMRA Key Accounts Manager, North America. TOMRA Recycling Sorting designs and manufactures sensor-based sorting technologies for the global recycling and waste management industry to transform resource recovery and create value in waste. The company was the first to develop advanced waste and metals sorting applications using high capacity near infrared (NIR) technology to extract the most value from resources and keep materials in a loop of use and reuse. To date, around 10,000 systems have been installed in 100 countries worldwide. For more information, visit www.tomra.com.