PI Vision in the research lab: how visualization drives insights at universities
“The price of an electric vehicle is really defined by the battery and the production of the battery cells,” Christoph Lienemann, Group Leader of Battery Production at PEM.In the previous blog, we looked at how large industrial enterprises are using PI Vision for operational intelligence and business optimization. Now, we'd like to examine how PI Vision helps researchers in the laboratory find new discoveries that open the door to the next technological breakthroughs.
The Science Behind Electrical Vehicle Batteries
In the realm of personal transport, fossil fuels are the established performers. For a long time, electrical vehicle (EV) batteries just couldn't keep up with petroleum. In 2014, personal electric vehicles were only 0.66% of new cars sold in Europe. But what if the goal of your research was to accelerate adoption of electrical vehicles by reducing the production costs of the EV battery, and to do that you had to replicate the manufacturing environment in your lab? That's exactly what the Production Engineering of E-Mobility (PEM) division of RTWH Aachen University set out to achieve.
Here's the story of how the Aachen's PEM division outperformed their expectations with PI Vision, a web-based visualization application for operational data.
Reducing Battery Costs by 20%
For a traditional internal combustion engine-powered automobile, the powertrain represents roughly 25 percent of the vehicle's total production cost. With substantially fewer parts than the combustion counterparts, the motors that propel an electric car represent only 15 percent of the production costs. However, the batteries to fuel those motors cost 36 percent of the car's value, and the combination of the two makes up a whopping 51 percent of an electric vehicle's total production cost!
“The price of an electric vehicle is really defined by the battery and the production of the battery cells,” explained Christoph Lienemann, Group Leader of Battery Production at PEM, at the 2017 OSIsoft Users Conference in London.
To meet the power and range demands travelers have come to expect, automotive transportation batteries must deliver a lot of energy from a compact, reusable cell. “The cell is a black box. Its chemical process is in there. You cannot open it. You can only destroy it and see what happened afterwards. It's a post-mortem analysis,” described Lienemann. Once a battery is produced, it must be energized, charged, drained, and then allowed to age for up to three weeks before final testing and QC verification can occur.
Meanwhile, each battery must sit idle with no reassurance the final product will be sellable, leaving a lot of capital resources sitting on the shelf. Past failure rates hover around 10 percent of total production.
With a goal of reducing end-of-line testing and failure rates, Aachen University organized a technology consortium to take a data-driven approach to the battery production process. The Aachen team knew there was untapped information available to them throughout the production line. “But now the challenge is to connect all this data, to connect the mixing process, to connect the coating process, to connect the welding process. Because it's such a different variety, you have electrical, chemical, and many influences. It's hard to understand what is really affecting the last 2 percent of the quality,” explained Lienemann.
A PI Vision display showing battery drying process overview.
In only six weeks, the consortium leveraged the PI System to capture data points across primary battery production processes. Using PI Vision, they easily visualized the data captured by the PI System to examine the cause-effect relationships between variables. Soon, Aachen was able to predict a reduction of production costs by 20 percent, while reducing scrap rates from 10 percent to two percent.
As the available data continued to increase, the Aachen team was be able to predict the final quality of an individual battery during production. Moving forward, Aachen plans to use the data to streamline, manage and direct the manufacturing process with an end-goal of reducing the scrap rate to below two percent, making electric vehicles far more affordable to the general public.
Get more information about PI Vision on our website and check out this quick demo.
To learn how PI Vision is used to manage research facilities at NASA and Harvard Medical School, read our customer success story here.