Data Analysis Companies Collaborate to Slash EV Battery Production Time
By detecting and diagnosing problems earlier in manufacturing, Voltaiq and PDF Solutions can reduce the number of defect-induced recalls, improve the yield of good cells and cut factory ramp-up time in half.
A battery intelligence pioneer will work with a venerable semiconductor yield-improvement firm in a partnership that promises to drastically accelerate the production ramp for the many new EV battery factories on the horizon. Voltaiq, the battery-analysis experts, and PDF Solutions announced the partnership today in Santa Clara, Ca.
Tal Sholklapper, Voltaiq’s CEO and cofounder, said the EV battery industry is in sore need of help shortening the manufacturing development cycle, which can take anywhere from four to 10 years from shovels in the ground to output of a consistent, quality product. “The automotive battery industry is really behind.” he said in an interview with SAE Media. “There is a lot of manual analysis and semi-empirical learning going on,” and that slows the discovery of future problems. He said the partnership had the potential to cut battery factory development time in half.
“The electrochemistry is new to [auto OEMs]. It’s a whole other beast,” he said, adding that there are millions of particles in a battery cell, and it doesn’t take many defective ones to cause poor performance or fires.
SAE Media also spoke with Sholklapper in January, when it was part of a group of journalists who saw a private presentation about Voltaiq’s efforts. “The complexity around batteries,” he said then, “is that some in the automotive industry like to think of them as mechanical and electrical engineers [would], so they'll think about them as electrical black box. Or, like every other mechanical component they have. The challenge is that these are electrochemical living and breathing organisms that also have those mechanical properties [expansion and contraction] you have to deal with and electrical properties you have to deal with. And you have to think about all of those in parallel.”
Leveraging already-existing data
That’s where Voltaiq’s Enterprise Battery Intelligence (EBI) platform comes in. It leverages thousands of data points from dozens to hundreds of cells that make up an EV battery, taking that data at any point of the production process and predicting future performance and behavior. The company works with firms in industries from transportation to consumer electronics and saw the need for its services in the full EV battery supply chain.
He said early on, the company studied two OEMs that seemed to be getting consistent yield and quality out of batteries being manufactured for them: Tesla and Apple.
PDF Solutions’ Extensio platform has modules that analyze data from manufacturing stations, process control, assembly and test operations. Michael Yu, VP of PDFS’ Advanced Solutions Group, said the Voltaiq partnership would speed its own entry in the battery market. “With more than 40,000 process-, test- and assembly tools and more than 15,000 engineers trained worldwide, our collaboration gets our best-in-class semiconductor process control methodology in the battery manufacturing arena.”
Both companies use AI and machine learning in their ultimate analyses. Sholklapper said that though both companies’ products are usable right now, an eventual goal is to put the combined data together in a single dashboard. He said that while analysis is useful at any point in the process, it’s critical during battery formation. That’s the process of initially running charging/discharging cycles on a battery cell, during which key chemical reactions take place that affect long-term energy density and reliability. The U.S. Department of Energy’s Office of Scientific and Technical Information says it typically takes many days or weeks and can create enormous bottlenecks in the manufacture of lithium-ion batteries in particular.
Of their combined efforts, Sholklapper offered a mind-blowing number: He said that one of the top three automakers recently did an analysis and concluded that by working with Voltaiq and PDF Solutions, they could perform 20,000 times the efficiency improvements versus what the OEM was doing manually.
Voltaiq CTO and cofounder Eli Leland stressed in January that this battery data already exists. “It already existed when we started the company in 2012,” he explained. “So, every cell phone has a log of its battery's performance. Over time, every vehicle, the battery management system is keeping the log in test labs where a lot of our customers are using our software. They've got huge facilities where they can test many thousands of batteries at a time — each independently — all that data is flowing into things like text files, with a time stamp, current and voltage. That data already existed on the manufacturing line. The last step of the manufacturing process is called formation cycling, where you're doing very tightly controlled charge and discharge cycles on a battery. That's the first time it produces that sort of electric, chemical heartbeat. All this data has already existed. We just realized that nobody was collecting it and putting it through really rigorous analysis.”
Last year, the company published the Voltaiq Data Format (VDF), a collection standard that helps smooth analysis efforts.
Analysis often a third-party endeavor
PDF Solutions, meanwhile, are masters of data coming from manufacturing equipment, other Internet of Things (IOT) devices and testing processes. They also have a 30-year history of training more than 15,000 engineers across the globe. They are experts in both metrology, the study and verification of measurements themselves, and data analysis. Sholklapper said PDF S has been proven to improve yield rates 50% faster than manufacturers can on their own.
In what may have been some foreshadowing, Sholklapper talked in January about why analysis-specific software markets get created for even some of the biggest verticals. “Some of our customers are Google, Amazon, Microsoft, and Meta, the largest software companies in the world. And they realize, you know, that it's vertical domain specific software ends up always having a market. You look at how there are tools for failure modes and effects analysis (FMEA), there's tools for exploratory data analysis (EDA), and none of these companies make them themselves,” the Voltaiq CEO said. “They used to make them as well, people could also rely on databases, they have smart software engineers, they could build their own database. And so especially around vertical specific software, you end up having very large multibillion dollar software industries that exists, it already exists in semiconductors, it's going to exist in batteries.”
Doing some back-of-envelope math, Sholklapper said the EV battery industry is going to need a lot of help to build and ramp up the 100 to 200 more battery gigafactories it will need if it is to supply an industry that hopes to get 100 million EVs on the road by 2030 to meet aggressive climate goals. And the industry has seen how much it can cost if production is delayed by mysterious quality problems.
A recent example is Ford’s choice to halt production of the F-150 Lightning in mid-February due to a battery fire. The company said a solution to the problem had been found and production would start again the week of March 13. Even at Ford’s low 2023 production goal of 55,000, that’s more than 1,000 trucks that won’t make their deliveries. Ford can likely absorb those costs. For a startup, such delays can risk the viability of the company.
The good news, Sholklapper said Wednesday, for future battery quality is that the automotive industry now “gets it,” and is actively seeking digital strategies for in-development and future gigafactories. He said the new partnership is quite scalable due to Voltaiq’s product being mostly software and PDF Solutions’ experience at putting boots on the ground at factories and training battery engineers to use the platforms.
INSIDERElectronics & Computers
University of Rochester Lab Creates New 'Reddmatter' Superconductivity Material...
INSIDERElectronics & Computers
MIT Report Finds US Lead in Advanced Computing is Almost Gone - Mobility...
Airbus Starts Testing Autonomous Landing, Taxi Assistance on A350 DragonFly...
Boeing to Develop Two New E-7 Variants for US Air Force - Mobility Engineering...
PAC-3 Missile Successfully Intercepts Cruise Missile Target - Mobility...
Air Force Pioneers the Future of Synthetic Jet Fuel - Mobility Engineering...
Driver-Monitoring: A New Era for Advancements in Sensor Technology
Manufacturing & Prototyping
Tailoring Additive Manufacturing to Your Needs: Strategies for...
How to Achieve Seamless Deployment of Level 3 Virtual ECUs for...
Electronics & Computers
Specifying Laser Modules for Optimized System Performance
The Power of Optical & Quantum Technology, Networking, &...
Electronics & Computers
Leveraging Machine Learning in CAE to Reduce Prototype Simulation and Testing
University of Rochester Lab Creates New 'Reddmatter' Superconductivity Material
INSIDERTest & Measurement
New Consortium to Develop Thermal Protection Materials for Hypersonic Vehicles
Multi-Agent RF Propagation Simulator
Low Distortion Titanium in Laser Powder Bed Fusion Systems
How to Test a Cognitive EW System