Software / AI

Advanced Analytics & AI in Battery Manufacturing The proliferation of sensors and inspection systems has shifted the challenge from data availability to data utilization. Advanced analytics and AI are shifting defect detection upstream, enabling earlier intervention, higher yields, and closed-loop process optimization across the battery value chain. CLOSED LOOP PIPELINE FOR ADVANCED ANALYTICS DATA ANALYTICS/AI DECISION ACTION FEEDBACK

Data Collection You need to collect all of these for advanced data analytics

IoT/FDC Equipment sensor data collection Process Metrology, run card, materials, MES, etc. Image X-ray, visible spectrum imaging, optical, etc. Testing End-of-line: formation, aging, QC and beyond

Data Lake -

Data Storage Data Characterization: Extract characteristics from raw information Data alignment and harmonization: Align data from electrode to stacked electrode to assembled cell

Governance & Trust: Model monitoring, data quality checks, version control, and auditability are expected in production AI systems.

AI & Analytics for Decision Support

Most issues are found at end of line or formation/aging. Insights from traditional and AI/ML data analytics enable manufacturers to identify problems earlier, saving time. This step includes AI model prep,training, & deployment.

Action & Control

Data insights now drive real or near real-time process adjustments, automated or operator-assisted, to prevent drift and stabilize yield.

Feedback & Monitoring

Continuous feedback ensures model accuracy, retraining, and compliance through monitoring and version control.

ANALYTICS APPLICATION EXAMPLES Overall Yield Dashboard Centralize data through yield trends, key parameters, and equipment performance

FDC Fault Detection & Classification) - Identify key control indicators Predictive Maintenance Minimize downtime across all equipment old and new

Source: Manufacturers Alliance, Siemens; MDPI, LinkedIn, Diligent, and all underlined content

Defect Management System DMS Connect all visual inspection to a central system and visualize defects along the electrode, layer, and cell

Advanced Process Control Use tool information and AI to, for example, direct thick electrode lanes to higher performing winders/stackers

Predict Quality Excursions Utilize AI/ML to recognize problems early on, before formation/aging Process Control & Tool to Tool Matching Determine out of spec tools, match tool performance, & identify root cause of tool issues Analytics for Formation/Aging Extract key parameters out of complicated formation/aging data

Extracted from the PDF; charts and images may read imperfectly.