Asked in episode 161 what has been hardest about building CATL's business in Europe, Matt Shen does not say regulation, permitting or energy prices. He says people. The process you can find. The equipment you can find. The people you cannot, because the industry has no history here.
His explanation for why the gap is people rather than paperwork is a distinction between mechanical engineering and chemistry. Mechanical is physical. Chemistry, he says, is something like it's alive: a great many conditions each make a small difference to the product, and knowing which ones and by how much is not a thing you read. His analogy is cooking. You have all the recipes, and you cannot cook exactly the same food.
That idea is the most contested one in a run of recent conversations. Several guests reach it independently, then draw opposite conclusions from it.
The word several guests reach for independently
CATL's first attempt at moving the knowledge to Germany was the obvious one: send locally hired staff to China for two or three months, then bring them back to run the line. It worked up to a point: too many things still came up that nobody knew how to handle. So CATL brought experts from China with two explicit jobs, keeping production going and teaching local colleagues how to handle the abnormal case. At the peak, 30 to 35% of the German workforce came from China. Two years later it was under 10%.
Yann Vincent, chief executive of ACC, arrives at the same place from the other direction in episode 160. Asked what a car person should expect on moving into batteries, he describes an industry that is a mix of mass production and craftsmanship. The chemistry side is the craftsmanship: you are implementing recipes, and when you change something you wait months to find out whether the product is still good. Nothing in the car industry works like that. The mechanical parts are different machines, but the way you drive and maintain them is familiar.
Celina Mikolajczak, Lyten's chief battery technology officer, uses the same word in episode 152 and means something different by it. Her subject is the workforce. Tolerances in cells are measured in microns and tact times at scale in microseconds, where a car plant works in fractions of a millimetre with people climbing onto the vehicle at each station. Nobody touches a cell. The usual conclusion, she says, is that if the work is not manual then workers are not needed, and she rejects it. Equipment pushing hundreds of thousands of cells a day needs people who know that at this position and that position the fine dust builds up a little every day, and know what to do about it. "This is craftsmanship," she says, meaning it as an argument about who gets to do the work: a route back to skilled blue-collar employment rather than a reason to expect less of it.
For Shen and Vincent, craft explains why the knowledge travels slowly. For Mikolajczak, it explains why the jobs are worth having.
Students, teachers, and who is asked to teach
If the knowledge is tacit, the awkward question is who Europe learns it from. Two European manufacturers answer the same way.
Vincent's third condition for closing the gap is the one that lands least comfortably, and he says it anyway: Europe should partner with Chinese cell manufacturers to accelerate its own learning, through joint ventures with IP transfer into Europe, majority held by the European company. His justification is a straight reversal. At one point we were the teachers and they were the students, and we have to acknowledge that now we are the students and they are the teachers. Pretending to catch up without their support would, in his word, be arrogant.
Sebastian Wolf, PowerCo's chief operating officer, puts the same posture in the first person in episode 147. The company needs, he says, "to not see ourselves as teachers, how Volkswagen and the engine world has been in the past". In cells it is the student, trying to learn and take the experience available now.
Wolf then attaches a limit that Vincent does not. Becoming a me-too producer will not help PowerCo, because something has to compensate for higher European labour costs, so success requires innovative products and innovative production technologies rather than a good copy. More than 50 production innovations sit in the funnel, dry coating among them. The implication is uncomfortable: a European cell maker that matches Asian process capability at European input costs still loses.
Julia Poliscanova of Transport & Environment supplies the policy version in episode 155. Yield improvement, she says, is not something a company reasons its way to. You hire people who went through it in China ten years ago and avoid repeating their mistakes. So she wants conditions attached to Chinese investment in Europe: local control of the venture, genuine IP and skills transfer, and a requirement to use local suppliers. On the Stellantis and CATL joint venture she is careful about blame. Nobody has asked the companies to transfer anything, so they do not.
Whether the Chinese price is craft, scale, or neither
The argument turns adversarial over what the Chinese cost advantage consists of, because that decides whether Europe can learn its way out of it.
Vincent will not quote a cell cost but will quote the gap: ACC is 20 to 25% more expensive than Chinese cell manufacturers. His reasons, in order. They started fifteen or twenty years ago and have gone down the learning curve. They were significantly subsidised on capex. Their energy is highly competitive. They work a lot. And they currently have overcapacity, pushing prices down further. Labour cost, he notes, is a minor part of it.
Vikram Handa, founder of Epsilon Advanced Materials, attacks the first item on that list in episode 159. The standard assumption is that Chinese graphite is cheap because of scale. He disagrees. Scale helps, but the prices quoted today are below marginal cost. The coke producer is probably losing money. So is the graphite producer. Most of these companies are publicly listed, so the numbers are there: everybody in the value chain is losing money and is all right with that, because the plants need to run. What you are competing against, on his reading, is not a cheaper production method but a price that reflects overcapacity and infrastructure that was provided rather than paid for. Mikolajczak reads headline prices the same way, as partly subsidised rather than real.
Poliscanova takes a third position. She dismisses cheap labour as an explanation that may have been true twenty years ago. The reason Chinese factories are cheaper today, she says, is very high automation and superior process efficiency, and Europe can copy that without paying anyone less. What remains is energy, around 50% more expensive than in China, low scale and yields because the industry is still learning, and missing vertical integration, with European companies getting completely different quotes from the same Chinese suppliers their competitors use.
What craft costs when the line is running
Mikolajczak's account of a coating line is the most physical description of why any of this is hard. A metre wide, moving at 50 metres a minute. Something goes wrong and it takes a minute to notice and react, and that minute is 50 metres of material to cut out. At a 10% scrap rate, she says, you cannot survive as a going business. In a gigafactory you can make a ton of scrap so fast that it costs a million dollars in a day. Somebody then has to decide, live, whether a machine scrapping a little more than normal justifies shutting the line down.
Frank Blome, PowerCo's chief executive, gives the same problem its balance-sheet form in episode 158. A cell plant running high scrap rates takes years to recover in margin. It goes into the balance sheet and, in his words, probably never comes back.
Asma Sharafi, who manages cell design at Cuberg, reaches the craft problem from the technology end in episode 135. Listing what defeated solid state for her, she puts above the chemistry problems one that is not chemistry at all: getting from lab scale to commercialisation, where defect densities appear and manufacturing complexity multiplies. Cuberg's design choice follows. It runs a lithium metal anode on a specialised liquid electrolyte rather than a solid one, so that it can use current lithium-ion equipment and the process knowledge the industry has built over ten or twenty years, which is the accumulated craft Shen says cannot be written down.
Mikolajczak's advice to anyone entering the field is flatter still: if you are making cells, you cannot get that education from papers or conferences. You have to be inside the building. Shen, asked how to support more leaders in the industry, does not suggest a training programme either. Send them to the plant for two weeks or a month, to see how a battery is actually made.
Where the humility argument runs out
Vincent, who wants Europe to sit at the Chinese feet on process, does not believe there are meaningful economies of scale in cell manufacturing. The process is, in his words, completely inflexible, and the more you invest the higher the risk. You get better equipment prices and amortise engineering across more product, and beyond that, not much. His observation on the largest announcements in the sector is delivered mildly: people were so excited about building the biggest factories ever.
Poliscanova explains why the European version is harder than it looks. Over 50% of battery investment decisions in Europe are being taken by European companies, and those companies are startups without manufacturing expertise, where in the United States it is mostly experienced tier-one South Korean players who already know how to build a factory.
Which returns to the line Vincent gives when asked how he balances manufacturing against next-generation chemistry. After five years in the industry he has a strong conviction that to innovate, you first have to master your production. As long as you are not mastering it, everything related to innovation is PowerPoint.