With the booming increase of low-volume and high-value-added production, injection molding manufacturing has become a pillar industry in China. In order to produce high quality product efficiently, it requires not only the process control but also the remote management, monitoring and planning. There are two key issues for remote management: how to collect and store the ever-growing process data and how to integrate and extend the current management techniques to the remote management system.

To deal with this large scale dataset, a cloud service structure for the injection molding industry is proposed in this work. This consists of a data storage service and a remote management service. It offers the business model of pay-as-you-grow to user. The storage service is based on Ha...[ Read more ]

With the booming increase of low-volume and high-value-added production, injection molding manufacturing has become a pillar industry in China. In order to produce high quality product efficiently, it requires not only the process control but also the remote management, monitoring and planning. There are two key issues for remote management: how to collect and store the ever-growing process data and how to integrate and extend the current management techniques to the remote management system.

To deal with this large scale dataset, a cloud service structure for the injection molding industry is proposed in this work. This consists of a data storage service and a remote management service. It offers the business model of pay-as-you-grow to user. The storage service is based on Hadoop distributed file system and HBase database system, which process strong scalability to handle the storage capacity. Based on the characteristic of the injection molding data, the index and filter of the database system are customized to enhance the query efficiency. Furthermore there database parameters are tuned for the developed cluster hardware in order to optimize the I/O performance. By utilizing the storage service, a preliminary remote management service is proposed with functions of remote monitoring and process data downloading. It is developed on the web pages based on browser-structure model, which is easy and convenient to access and use.

The proposed cloud service has been tested and verified to a laboratory-scale cluster and injection molding controllers. The experimental results have demonstrated the effectiveness of the proposed service. The successful completion of this study not only addresses these important practical application issues, but also provides a cloud service suitable for the characteristics of injection molding.