New Shuffle Conveyor Featuring an Innovative Drive System
Practical Technology for Unmanned Distribution Center Applications
Tsubakimoto Chain Co. (Headquarters: Kita-ku, Osaka, President: Yasushi Ohara) has developed a globally unrivalled, rapid, high-precision image recognition technology that makes optimal use of artificial intelligence (AI) as a key technology for unmanned distribution centers. Going forward, we intend to make this technology practical by conducting trial experiments for customers, with sales scheduled to start from fiscal 2021 onward.
Overview of technology
Our company develops and delivers materials handling systems used in a range of industries for transporting, storing, and sorting goods. One such industry is logistics, where the recent expansion of e-commerce has underlined the need to contend with labor shortages and calls for shorter working hours. Labor saving and unmanned operation is required throughout the process, including receiving, inspection, stock management, inventory, and packing. The need to further streamline warehouse picking is particularly urgent.
In response, our company has taken advantage of image recognition technology to detect the shape, color, and pattern of products for sorting, particularly for unmanned and non-barcode operations, and developed AI image recognition technology capable of determining key features rapidly with high accuracy. We are collaborating with the AI startup business EAGLYS Inc., with our product development prowess and technical insights, using the EAGLYS AI algorithm’s design and secure computation technology, in trial experiments centering on learning and inference. In the process, we successfully enhanced recognition accuracy by reducing image data noise while tapping into original deep learning. We also devised our own AI algorithm with ultra-rapid response time (less than 0.15 seconds), high recognition rate (98.55%), and high accuracy (99.99%) in product recognition. Our rapid, high-precision AI image recognition technology is unmatched around the world.
Going forward, we intend to commercialize this technology for use in unmanned distribution centers and to expand the range of possible applications.
Head office: Shibuya-ku, Tokyo/President and Representative Director: Hiroki Imabayashi. Founded in December 2016. A young and lively startup that has established partnerships with companies in various industries under the slogan "Constructing highly accurate and practical AI systems" by leveraging core technologies of AI algorithm design and secure computation.
Received the Small and Medium Enterprise Agency Commissioner's Award at the Japan Venture Awards (JVA) 2020. Following in-depth assessment of the top 191 venture companies, in areas including business content, activity results, and managerial qualities, the business was highly commended for outstanding management.
Rather than having to enter barcodes for various products of different shapes and sizes (miscellaneous daily staples, food, commercial drugs, etc.), product information can now be determined rapidly and accurately, simply by capturing product images. This paves the way for next-stage automated transport processing and unmanned operation. Rolling out this technology can help eliminate hindrances such as the need to enter barcodes manually and automatic barcode reading errors, which will further streamline work and expand the potential for unmanned distribution centers.
Plans are already being developed in areas where barcode use is not feasible, including differentiating production factory parts, assessing when agricultural produce is ready for harvesting, and picking assistance for building components.
Unmanned Distribution Center example