ZPower Recharges Production with TIBCO Data Science
Fine-grained control over processes, reduced scrap, reliable profitability
ZPower needed a process control system to get to higher volumes without sacrificing small-batch quality. The solution needed to encompass all manufacturing processes and provide visibility into all production steps, while delivering quality product to customers.
The company turned to TIBCO Statistica™ software (part of the TIBCO® Data Science platform) to address these challenges. "There is no package in the world more flexible," said John Coleman, senior data scientist. "If you have more data, real-time sensors, or multiple data streams, you get more out of it."
Fine-grained control over processes
With over 20,000 streams of information coming from the production line, visibility and control were critical. TIBCO Data Science enabled ZPower to apply a powerful multivariate analytic technique called "multiple stream control charts" for effective state-of-the-art process monitoring. This is the type of solution that the company couldn't simply buy from other vendors because it requires flexibility and scripting capabilities to craft the actual method of control.
"Statistica software allows us to tweak, twist, and bend analytic routines so they fit into our manufacturing process and, for example, let us monitor battery testing," said Coleman.
"The methodical use of structured problem-solving techniques, followed up with the systematic implementation of SPC using the Statistica component delivered to the factory floor via visual displays resulted in higher quality, higher yield, and lower cost product," said Chris Fredric, director of process engineering.
Reduced scrap, reliable profitability
ZPower also improved process yield by reducing scrap. "The internal yield improved significantly and continues to improve," says Coleman. This achievement is critical because it means the company can have the appropriate amount of material for consistently producing the volumes necessary to satisfy growing customer demand. In addition, its production lines can move more efficiently.
Augmented intelligence, better decision-making
Use of TIBCO Data Science has spread across the factory; dashboards are everywhere. Everyone from engineers to data scientists are using the tools. On the manufacturing floor, process stakeholders and owners of all skill levels drill into detailed reports, graphs, and analyses. Simple and descriptive dashboards provide 360-degree views and include colored traffic lights that efficiently communicate the status of complex, highly dimensional process streams. These traffic light displays can be configured by engineers to report on workflows, analyses, and processes to drive real-time decision-making.
"The analytics journey continues well beyond just control charts and stabilizing processes," said Coleman. "The natural next step is to optimize those processes."