Integrating Automation and Big data in Lithium-ion Battery Manufacturing: A case study of Ultium Cells Joint Venture

By Joonghyun Song

In recent manufacturing business, integrating automation and big data is considered as mandatory investment. In overall manufacturing business, automation rate is estimated to be approximately 20-30%. In highly automated industry such as automotive industry, the rate can be over 70%. In battery manufacturing, some batteries are still manufactured with extensive manual labor. However, as battery application expands to the industries where the product can directly impact on customers’ safety, quality assurance is the largest factor of automation and big data infrastructure; these industries include Electric Vehicle (EV) and Energy Storage System (ESS).

For successful integration of automation and big data in Lithium-ion battery manufacturing, several requirements need to be satisfied. As Lithium-ion battery manufacturing is recently growing, experienced engineers are insufficient; these engineers can correctly operate automation and interpret big data based on their battery manufacturing background. Another requirement is to design reliable network system. Data communication of automation and big data is established on the network system. Providing adequate storage and safety back-up in case of network failure or blackout will ensure stable manufacturing. It is true that automation and big data installation require higher cost and longer stabilization period, but increase in productivity, production efficiency, and quality will realize quicker return of investment.

Lithium-ion battery manufacturing facilities require large factory footprint. For instance, Ultium Cells LLC in Ohio is 2.8 million square feet. As multibillion-dollar battery manufacturing facilities are under construction competingly across the world, delays in mass production will result losing market share. These facilities have sophisticated production parameters and create millions of production data sets every day. Manufacturing Execution System (MES) is an effective tool to achieve process stabilization. The automated system is designed to control all process steps, collect real-time production data, and trace each product. Hence, it can quickly identify ramp up bottleneck and allocate manufacturing resources where needed. Implementing MES will bring mass production plan forward.

After successful start of mass production, maintaining operational performance is important. As Lithium-ion manufacturing process is complicated, chances of process variability are always present. These may be beyond MES’ process control such as ambient condition change, time delay, and machine failure. In order to monitor these contingencies, Statistical Process Control (SPC) is introduced. The system creates statistical report and determines if a process is out control limit based on manufacturing data from MES. The system is able to send warning via mail or text before quality risk aggravates. Furthermore, SPC is often integrated with control room to assist operation monitoring. All these efforts contribute to consistent manufacturing.

Automation and big data can achieve high standard of quality assurance. The automation systems can achieve manufacturing consistency, which promises uniform battery quality. They are also effective when process control fails. Analyzing real-time production data identifies the source and root cause of process control failure. It reduces production loss due to production stop and producing more defects. Moreover, traceability of each product allows to obtain accurate risk range. As a result, implementing automation and big data will minimize quality cost and retain strong trust from customers.

Case Study 1: Automating logistics system
Ultium Cells is a large battery manufacturing facility with annual capacity of 40GWh, which is equivalent to 400,000 EVs. In order to fully operate the facility, incoming raw material, in-house semi-product, and outgoing finished product must move continuously to the right place at the right time. In Ultium Cells, the logistics system is fully automated by using combination of Automated Guided Vehicle (AGV) and conveyor system. MES executes moving command based on the route information it has. The automated system also monitors the current location of each product in case of command failure or process delays. Moreover, control room displays all logistics information and provides real-time 3D map to assist process control.

Case Study 2: Implementing vision system
Development of camera and image processing method has allowed broader application to manufacturing. It is more accurate, consistent, and faster than human eyes. Ultium Cells applied vision system extensively to different processes across the plant. The vision system is integrated with quality criteria to reject dimension and cosmetic defects. Production engineers have access to all image data, which are analyzed to prevent process deviation. Furthermore, implementing interdependencies with previous processes allows the system to easily identify the source of defects. The image data is also useful for quality engineers to confirm the risk range.

An emerging technology to develop automation and big data even further is Artificial Intelligence (AI). Automation and big data still rely on data engineers’ analytics, and an incorrect interpretation may lead to wrong decisions. AI implementation to the manufacturing system allows production to be more predictive. For instance, AI can predict machine failures by analyzing machine data, minimizing production loss. Another advantage of applying AI, especially to Lithium-ion battery manufacturing, is that it can predict battery performance. Current technology is limited to predict battery performance from material changes, so engineers are dependent on test and sampling statistics. Ultimately, AI-driven automation will design and optimize production process on its own analytic decision.

Automation and big data have contributed to Lithium-ion battery manufacturing by enhancing productivity, production efficiency, and quality assurance. The automated system has reduced the cost of battery manufacturing to compete with Internal Combustion Engine (ICE). Most importantly, consistent battery quality significantly improved customers’ concern over safety. Overall, integration of automation and big data allowed expansion of Lithium-ion battery to transportation and provided customers clean alternative to ICE vehicle. Further development of automation and big data by integrating with emerging technologies, transition from ICE vehicle to EV will be successful, leading to further expansion of battery applications.