Make a substantial investment, set up battery production, and then reap huge profits. Numerous examples from the recent past have shown that this is not how it works. One key sticking point is quality assurance. “The critical characteristics in battery production are well known – and yet the difficulty of complying with the required tolerances is often underestimated,” says Dr. Philipp Jatzkowski, Senior Manager Quality Assurance Consulting at Testo Industrial Services. In an exclusive interview with Battery-News, the testing process expert explains what is important now and will be in the future.
Battery-News: What are the biggest obstacles in quality processes in the field of battery cell manufacturing?
Dr. Philipp Jatzkowski: The challenges in battery cell production result from the complexity of the product, the production system, and the tight product and process tolerances. Studies show an average scrap rate of 30 percent during production ramp-up. Reducing this rate is currently a very slow process, taking several years. Even the smallest deviations can add up along the value chain and lead to high rejection rates. Many European manufacturers still lack detailed experience in production ramp-up and in scaling such production systems.
“Reliable data on the capability of the manufacturing process are essential to ensure stable and reproducible production procedures.”
Battery-News: At which points is quality assurance particularly critical to success in cell production?
Jatzkowski: Nearly all production steps in cell manufacturing place high demands on production systems and measurement technology. Therefore, it is not a single process that is particularly critical. Rather, a consistent quality assurance concept is required throughout the entire value chain. This includes the systematic establishment of control loops with the targeted use of inline measurement technology, production-related tests, and laboratory analyses. This requires calibrated measurement systems with proven measurement process suitability – often referred to as “measurement system analysis” or “MSA.” In addition, reliable data on the capability of the manufacturing process are essential to ensure stable and reproducible production procedures. With the help of statistical process control, it is possible to detect deviations at an early stage and avoid scrap in the long term.
Battery-News: What are typical mistakes made in systematic measurement and test equipment planning in the field of battery production?
Jatzkowski: A common mistake is to consider individual sensors and measuring systems in isolation, rather than planning them as an integral part of a multi-stage quality assurance concept with defined control loops. Test planning is complex and requires all relevant influencing factors to be taken into account. In this context, we also refer to measuring system, man, environment, method, and measuring object – supplemented by the measuring frequency, which is decisive for statistical process control. Test planning must be carried out by metrologically trained specialists and concluded with proof of the measuring processʼs suitability.
“A robust starting position and growing experience with the product and production system are crucial.”
Battery-News: The industry is developing rapidly. How is it possible to establish reliable testing processes when new technologies are constantly emerging?
Jatzkowski: To do this, the characteristics that determine the quality of the end product must be identified during the development process, and stable testing as well as manufacturing processes must then be established for precisely these characteristics. On this basis, continuous improvement can be driven forward – for example, through optimized cell design, more efficient production systems, or the use of new measurement technology such as high-resolution machine vision systems with machine learning algorithms. Changes always take place gradually. A robust starting position and growing experience with the product and production system are crucial.
“We need to gain a better understanding of when a deviation in the coating of electrodes is relevant to safety or performance.”
Battery-News: Traceability is of central importance in battery cell production. How can we ensure that test and quality data is consistently recorded and usable throughout the entire process chain?
Jatzkowski: I donʼt see traceability as much of a challenge, because product identification systems, such as QR codes, and automated data storage already enable digital cell records today. The more exciting question is how we use these data to deepen our understanding of the product and production and to control quality. Artificial intelligence will play a central role in this. If we generate high-quality data with low uncertainty and complete meta-information, we create a valuable treasure trove of information. Itʼs crucial what questions we ask of these data. Big data and AI are not ends in themselves! The goal must be to identify functional specification limits across multiple process steps and to design a robust production system. One example of this is the detection of defects in electrodes. We need to gain a better understanding of when a deviation in the coating is relevant to safety or performance – and how such deviations arise along the process chain, from slurry mixing to electrode cutting.
“As output increases, a database is created for the continuous improvement of product and process quality.”
Battery-News: Many production lines are currently being set up or ramped up. How can quality assurance be effectively implemented when stable series production is not yet underway?
Jatzkowski: Every ramp-up process is iterative and involves a steep learning curve – even if comparable production lines are already in operation. Itʼs about setting up, commissioning, and interlinking technically sophisticated production systems. This process requires very close and trusting collaboration between the cell manufacturer and the supplier of the production system. In the ramp-up process, we go through defined phases with a continuous increase in the stability and output of the production system. This makes it all the more important to ensure the suitability of the measurement and testing processes as well as the capability of the production systems right from the start – even if this initially only involves preliminary machine performance based on low quantities. As output increases, a database is created for the continuous improvement of product and process quality.
Battery-News: Which physical or process-related parameters are often underestimated but are crucial for product quality or safety?
Jatzkowski: The critical characteristics in battery production are well known. Nonetheless, the difficulty of complaning with the required tolerances is often underestimated. In electrode coating, these tolerances are often just a few micrometers. For the measuring system, this means a stable measurement uncertainty of less than one micrometer must be maintained. By way of comparison, bacteria are one micrometer to three micrometers in size! In contrast, most electrical measurements – such as the voltage at the battery cell or the insulation resistance – are not particularly challenging. In my experience, we have a good balance between measurement uncertainty and tolerance here.
“Often, the challenges lie in understanding fundamental terms.”
Battery-News: How closely should quality assurance, production IT, and data analysis work together today? Is quality assurance increasingly becoming a digital issue?
Jatzkowski: Collaboration at the expert level is essential. This starts with communication between product design, production process planning, and quality assurance. One of the biggest problems is that there is too little communication in this area. A crucial prerequisite for a stable production system is that all parties speak the same language and pursue the same goal – namely a production process in which compliance with functional tolerances is ensured with sufficiently low uncertainty via stable control loops. However, the challenges often lie in understanding fundamental terms such as “functional tolerance”, “machine and process variation”, “measurement uncertainty”, “acceptance limits”, or “intervention limits and error slip” – all of which should be part of the production control plan. This circle is then expanded by experts from production IT and data analysis, who create the basis for reliable data acquisition and storage and provide the information required to use the control loops for quality assurance and pattern recognition for continuous improvement.
“We must boldly use digital technologies and understand them – including their limitations.”
Battery-News: How is digitalization changing quality assurance, for example through AI-supported testing procedures?
Jatzkowski: The change is fundamental. AI-supported processes are becoming increasingly important. We must boldly use such technologies and understand them – including their limitations. A machine learning model based on a neural network is not automatically better than an analytical model. In many cases, classic analytical models are faster and more accurate. Therefore, we must always check what prior knowledge we can bring to our control loops. It is also possible to combine analytical models with neural networks where modeling is too complex or too costly. We must not underestimate the training effort required for AI models. The quality of the input data is just as crucial. Sensors and measurement systems must provide reliable, traceable data with low uncertainty. There is no point in combining a machine learning algorithm with a machine vision system that delivers blurred images. The same applies to AI-based testing processes. The suitability of the measurement process must be verified. To do this, the algorithm has to be frozen and thus tested in a specified release. It is permissible to continue training the algorithm in the background. However, as soon as it is to be used productively, its suitability must be proven again.
Battery-News: In the race for production capacity, factories are being built in record time. How can we ensure that quality does not become a secondary concern?
Jatzkowski: The answer is simple: It will only be possible to build production capacity quickly and sustainably if quality assurance is understood as an integral part of the production process from the outset. Thatʼs nothing new. It forms the core of the APQP approach – Advanced Product and Quality Planning. This concept is based on the consistent application of methods for verifying measurement process suitability and manufacturing process capability. It has been known since the 1980s and has proven itself many times over. Unfortunately, the approach is not consistently implemented in many companies.
“Quality assurance is an integral part of value creation. This must be the managementʼs goal and has to be exemplified every day.”
Battery-News: What are the most common cultural or organizational obstabcles when setting up a functioning quality assurance system – and how can they be overcome?
Jatzkowski: A stable and capable production system is created when at least three parties work closely together: product design, production process planning, and quality assurance. This is the only way to design economical production systems in which deviations in the manufacturing process, measurement uncertainty, and tolerances are coordinated in such a way that production interruptions and scrap are minimized. Clear understanding and active support from management are just as crucial. Quality assurance is an integral part of value creation. This must be the managementʼs goal and has to be exemplified every day. Clear processes, defined competency profiles, and appropriately trained employees are required. Only when everyone knows exactly what their responsibilities are, what tools they can use, and what standards they must adhere to can quality be reliably implemented in day-to-day business.
“We must achieve our goal of establishing stable cell production in Europe within five years at the latest.”
Battery-News: Looking five years into the future, how will quality management in cell production develop, and what role will companies such as Testo Industrial Services play in this?
Jatzkowski: We must achieve our goal of establishing stable cell production in Europe within five years at the latest. The European automotive industry needs its own high-performance production capacities. Quality will be a key differentiator. The “Quality Made in Europe” label will become a promise for durable, high-quality, and sustainably produced batteries. To achieve this, production capacities must be greatly expanded. At the same time, manufacturers need to build up practical experience in cell design and production. Europe still lags behind Asia in this respect. But Iʼm confident that we will catch up. We have strong machine and plant manufacturers, leading research institutions, and a deep understanding of quality and technology. If we combine our production expertise with European quality standards, we can establish competitive and independent cell manufacturing in Europe. At Testo Industrial Services, we see our task as laying the foundations for quality assurance – through accredited calibrated measurement technology, systematic test planning, and the consistent development of quality assurance strategies. Specifically, we provide support in planning suitable test procedures, verifying the suitability of measurement processes, evaluating production systems, and establishing stable control loops – in line with statistical process control.
