Predictive IoT Solution Case Study: Reaction Injection Molding Equipment
A manufacturer of high-performance aftermarket automotive parts was having problems with its two-part reaction-injection molding (RIM) mixing nozzles clogging. The situation led to excessive costs in increased downtime, waste and labor.
To overcome these challenges, a custom built hosted software solution was used to predict nozzle clogging in real-time and alert operators when needed and before the clogging occurred. The IoT solution also remotely monitored other criteria and was able to help identify the root cause of issues affecting quality.
With the introduction of the predictive IoT solution, clogs were eliminated, fewer nozzle changes were experienced and part quality improved, reducing scrap. The predictive software solution led to insights that reduced raw material consumption by increasing the time between material purges and reducing purge volumes when a purge is needed. Cost savings was approximated at $450,000 annually with about a 7% increase in machine performance. The manufacturer achieved a return on investment in less than one month.
Traditional approaches to machine learning require a large number of sensors collecting data for a long period of time, so that the machine learning algorithm can eventually learn how to optimize machine performance. In contrast to this, the simulation model was able to predict nozzle clogging and provided immediate value creation with relatively few sensors. Contract to completion took only 4-5 weeks including integration.
Predictive Maintenance Solutions
RRAMAC has been providing predictive maintenance solutions for over a decade, covering a broad range of industries and applications. The three basic techniques for implementing predictive maintenance are condition monitoring, machine learning and simulation. The simulation approach to predictive maintenance leverages human knowledge of cause and effect relationships which apply to the specific machine or application. These relationships are applied to simulation software which then predicts failures.
RRAMAC is working with industry experts across a broad range of industries to provide simulation base predictive maintenance solutions to our customers. These systems can be hosted on RRAMAC’s cloud servers or configured as part of an on-premise hosted solution. The simulation software runs on servers, on edge nodes, or a combination of both, depending on the speed of the application. To learn more about increasing machine performance and other IoT solutions provided by RRAMAC and our team of resources, contact us today.