'Self Learning' Sorting Machine for Battery Recycler in the Midlands

11 April 2012 UK waste battery collection and recycling company, G & P Batteries is investing in an automated battery sorting system from Optisort that utilises a neural network to 'self learn'. According to Darlaston based G & P, the new equipment was developed by Optisort, and is capable of sorting at a rate of more than five batteries a second. More Waste Management World Articles Rise of the Machines: Robot Recycling 'Self Learning' Sorting Machine for Battery Recycler in the Midlands Negative to Positive Sorting Reaching for the stars with SATURN High Resolution NIR Sorting System for Multiple Waste Streams from MSS Black bags in, commercial grade recyclate out Innovative sorting: an essential for economic improvements in waste handling Case Study: Eriez Upgrades Scrap Drums for U.S. Metal Recycler OmniSource Construction Waste - Hong Kong Style Installation is due to take place later this year and once commissioned, the company said that the equipment will more than double the sorting capacity of portable waste batteries handled at its site in Darlaston, West Midlands. The Optisort system is designed to recognise most common brands and types of batteries and so speeds up the sorting process, which the company said will enable it to easily handle the uplift in volumes necessary for the UK to meet its battery recycling obligations under the UK and EU targets. The firm added that it will always require the expertise of manual sorters, whose experience enables them to recognise accurately the lesser known battery brands and chemistries. G & P said that 2012 is a crucial year for the UK, which is tasked with meeting its first significant battery recycling target under the European Directive. By the end of the year, the country will need to have collected 25% of the volume of batteries placed on the market. Last year's target was 18%, and results issued by the Environment Agency so far show that it is likely this will be achieved. Neural networks: learning capabilities According to the Swedish manufacturer, the modular Optisort Battery Sorter can automatically sort between 4 and 8 tonnes of waste portable batteries per day (0.5 to 1 tonne per hour) and can handle all types of portable batteries from D to button cells (button cells are separated but not sorted), including small cell phone batteries. In addition, the company said that all chemistries available today can be sorted and it's also possible to add new sorting fractions, such as different kinds of Lithium batteries. The machine identifies the batteries by utilising a vision sensor that captures images of every battery in high speed. According to Optisort, the system then uses a neural network to compare the image to previous images taken in a large set of data in milliseconds. After classification the batteries are separated with airjets into each respective fractions which the manufacturer said allows for high accuracy under high speed. The company also claimed that the system is self learning, which enables it to recognise batteries even if they are dirty or damaged. In addition to the actual sorting of batteries the system collects data on the sorted batteries such as brand, model, size and chemistry. Optisort said that this enables detailed statistics to be extracted which then can be used for reports and analysis. For compliance schemes the statistics may be used to find freeriders but potentially also to provide producers with valuable market information. According to Optisort the system always provies fractions that are at least 98% pure and is also possible to sort out known counterfeit batteries and also to identify new ones. Read More Rise of the Machines: Robot Recycling Recycling robots that think for themselves. Automated arms that can learn and get hurt. Is this something out of a science fiction novel? Nope, but a Finnish solution currently on trial sorting construction and demolition waste with SITA Finland. Tom Freyberg investigates. New C&I Recycling Facility in Exeter to Use NIR Sorting A new Commercial and Industrial Material Recycling Facility in the Exeter, UK has selected sensor based sorting equipment specialist, TITECH as its technology provider of choice. Prototype to Recycle Human Waste into Energy & Water A team of researchers from three universities in the UK are undertaking a project to build a prototype of a device to harvest energy and clean drinking water from human waste Free Magazine Subscription Free Email Newsletter