Lithium-ion battery (LIB) market size will grow at the compounded annual growth rate of +16%, surpassing 165 billion USD in revenue by 2030. While most of this growth is enabled by innovations in battery chemistry and performance, safety hazards remain a primary concern and a major restrain to market expansion. Safety incidences often results from errors introduced during high-volume manufacturing that, through a chain of events, leads to a thermal runaway and fire. C4V has teamed with leading Machine Learning (ML) and Artificial Intelligence (AI) experts to address the issue by developing tools that can not only track and minimize or eliminate manufacturing defects in the battery cells, but can warn users of an incipient catastrophic event ahead of time, thus preventing any damage to property or loss of life. The first generation of the DigitalDNA (DDNA) software is able to automatically capture key electrochemical data from cell cyclers installed at iM3NY, a New York based gigafactory that produces 50Ah prismatic cells.[1] DDNA automatically curates and analyse data generated at the production floor, and create actionable outputs for operator to take corrective measures in near real-time. DDNA is designed to be platform agnostic and can capture data in multiple formats. The next-generation of DDNA, with an in-built advanced data analytics algorithms can easily access and use these large data sets to train the ML models and implement AI to provide predictive insights and enable continuous improvements in electrochemical performance of LIBs. In addition, with a recent release of the Supply Chain module, DDNA can perform a full inventory control and management of +30 components that are required for battery cell production. With this module, DDNA can also track the progress of C4V’s extensive raw material qualification program currently underway for more than 50 vendors globally. By integrating best-practices in laboratory information and data management systems, DDNA enables a high level of information flow control during the 5 discrete stages of phase-gate qualification process starting with a preliminary  assessments in a coin cell to a full scale evaluation in the commercial cell. When integrated with the warehouse management system and high-level business systems such as ERP, the predictive capability of the software can use the raw material utilization data intelligently to create sourcing and procurement scenarios to achieve full inventory and cost optimization. 

By leveraging the fully digitized future gigafactories and the IIoT ecosystems, the future-generations of DDNA will seamlessly integrate with manufacturing execution systems to collect data at each step of the manufacturing process. By tapping into the real-time visualization of process and equipment performance data, the ML and AI analytics will be able to detect anomalies ahead of time. This ability to predict failure and perform preventive maintenance will increase equipment availability and performance and reduce disruptions and costly repairs. More importantly, DDNA will interface with the numerous in- or at-line quality control instruments implemented in a roll-to-roll process. Together with feedback control loop, access to statistically significant anomaly data set and advanced descriptive analytics, DDNA will enable processes that will reduce waste and improve yields. By quickly detecting and eliminating any debilitating defects at every step, DDNA will afford a highly reliable and safe battery products. 

We envision that DDNA will mature into a comprehensive software platform that will enable Smart gigafactories and predictive manufacturing and will make intelligent decisions informed by data gathered over the entire value chain of LIB from molecules (mines) to machines (vehicles and Energy Storage Systems).