■ Background and outline of system development In recent years, fires caused by lithium-ion batteries installed in small electronic devices have been increasing in crushers of waste treatment facilities. The moment these lithium-ion batteries are crushed by a crusher, they become sparks and a source of high-temperature heat, which ignites the surrounding waste plastic and causes a fire. The general remedy is to install the smoke and heat detection system on the conveyor belt after crushing, but with this method, the fire spreads late because the fire has already occurred when it is detected on the conveyor. It means detection at a very high risk. In addition, when a fire actually broke out, it was difficult to quickly extinguish the fire because the fire type moved on the conveyor and it was difficult to locate the fire type due to smoke. Therefore, EAI Co., Ltd., which owns AI / IoT technology in the environmental field, and Daiei Environmental Group, which is developing a resource recycling business, have jointly developed an AI spark detection system "Spark Eye" to solve this problem. bottom. This system is a system in which AI automatically detects sparks immediately after crushing based on the camera image installed around the crusher and issues an alert (automatic watering cooperation is possible). This makes it possible to detect a fire risk at a stage earlier than before and prevent a fire in advance. In addition, it is equipped with sensitivity adjustment and visualization functions (detection number graph, spark image, video storage for 20 days or more, real-time monitoring function, etc.) by setting the alarm level, which enables quantitative evaluation of fire risk. In the future, both companies will aim for safe and secure waste treatment by expanding these systems to local governments.
(Left) Moment of crushing lithium-ion battery (Right) Spark detection by the system
■ System overview and configuration
■ Main functions of the system
■ Scheduled to start selling around December 2020 ■ Delivery time Approximately 1 to 2 months after receiving an order * Peripheral work (watering piping, etc.) is not included ■ Product page https://www.sparkeye.net/
