Feasible Inc. is developing a battery inspection platform, called EchoStat^®, that uses ultrasound and data analytics to scan battery cells rapidly and non-invasively, and deliver spatially-resolved insights about internal structure of cells. The company is focused on improving battery manufacturing quality, yield, and consistency, thereby decreasing battery cost and reducing inefficiencies in utilization and system design.

Feasible’s NSF SBIR Phase I award (# 1621926) mainly focused on designing, assembling, and de-risking the EchoStat platform’s hardware and electronics to enable repeatable, faster, and high SNR measurements on battery cells. Feasible also improved data collection and analysis pipeline infrastructure to be robust and secure. 

Feasible’s NSF SBIR Phase II and IIB award (# 1831080) were used to demonstrate the value of EchoStat inspection during battery manufacturing cell finishing (with particular emphasis on electrolyte saturation quality) and end-of-line quality inspection. The finishing process, on average, contributes ~30% of the overall yield loss and >70% of total production time in lithium ion battery (LIB) cell manufacturing. Cell wetting is a critical step where insufficiently soaked regions of the porous electrodes can result in poor, irreversible passivation (SEI) layer formation, electrolyte breakdown and lithium plating; all of which reduce battery performance and lead to safety issues. EchoStat is sensitive to subtle changes in a battery’s physical integrity and construction quality. By collecting data at multiple positions across a cell, EchoStat delivers spatially resolved information in seconds, a key advantage as electrical methods only deliver cell-average insights.

Monitoring electrolyte distribution - One high-value application in cell production is monitoring of the electrolyte wetting process. Ensuring proper electrolyte distribution is crucial for safe, reliable performance, especially as cells get larger and more energy dense for EVs and ESS. However, there is currently no direct, non-invasive way to inspect this process at scale, so while there is significant variability in production, with standard methods it is difficult to correct for this as detection can occur weeks after errors occur. As ultrasound is well-suited to detect the distribution of fluids in porous media (i.e., most of the components in a Li-ion cell), this presents a unique opportunity for impact. Because of its ability to quickly assess variation of internal properties, EchoStat can deliver information about the effect of process parameters on electrolyte distribution and is sensitive to changes in electrolyte amount. Multi-position EchoStat measurements are used to visualize and quantify electrolyte saturation behavior in cells. EchoStat measurements during soaking can provide value in process development as well as in production. Aggregated EchoStat metrics from spatially resolved measurements on cells at the beginning-of-life correlate better to cycle performance than formation capacity, voltage, and other electrical methods. With support from NSF SBIR Phase IIB, Feasible worked with a European EV maker at their cell pilot facility to shows how EchoStat could accelerate process development, improve yield, and predict cycle life performance by detecting process variation earlier and with greater sensitivity than state for the art electrical methods.

Defect Detection – With a growing demand for larger and more energy dense cells, the probability of occurrence of manufacturing defects has gone up, while the ability to detect small but critical defects with cell-averaged electrical methods has decreased. Detection of manufacturing defects, like tears and folds in electrodes or separator, improper tab welding, as well as metallic inclusions, is crucial for safe, reliable, fast-charging EVs.

These technical results were presented to a number of potential customers and development partners, including battery manufacturing majors LG Chem and Samsung SDI. Feasible also won LG Chem’s 2019 Battery Challenge. In 2019, Feasible learnt that LG Chem is interested in the detection of separator folds and misalignments that occur during cell assembly. These defects are the suspected cause for the $2.2B recall of the GM Bolt EVs over the last 2 years. One of the main challenges of detecting manufacturing defects like separator (very thin and flexible) folds and misalignment, as well as other defects like metal particle inclusions, is that they are very difficult to detect with optical and X-ray inspection methods, and most often do not show up till months after the cells are operating in the field resulting in poor battery performance, or at worst, catastrophic fires. 

There is a lot of interest in implementing non-invasive and spatially-resolved inspection solutions for detecting these manufacturing defect before battery cells leave the factory. Feasible is in discussions with various EV and battery makers for EchoStat product validation and evaluation projects. Beyond these evaluation projects, Feasible plans to demonstrate EchoStat’s in delivering insights at production-scales to decreases in production costs and process times alongside improvements in quality and performance.

Last Modified: 08/26/2021
Modified by: Andrew G Hsieh