I’ve recently watched a video by Patrik Boyle, ”The Inevitable Decline of WeWork”, which ended with the story of Adam Neuman, the founder, raising another $ 300 M+ after crashing WeWork, on some unspecified startup in, ahem, real estate. I had a good laugh until my LinkedIn feed supplied me with an announcement in a timely manner that Mr. Peter Carlsson, the ex-founder of Northvolt, had launched another venture, Aris Machina AB. Source: https://www.bloomberg.com/news/articles/2025-05-09/ex-northvolt-ceo-gets-new-funds-for-ai-startup-in-manufacturing?utm_source=website&utm_medium=share&utm_campaign=linkedin?sref=3gAbaqte The venture would use artificial intelligence to optimize manufacturing processes, and according to co-founder Siddharth Khullar, it has already attracted early-stage investment from investors such as Earlybird, Village Global, AENU, and Planet A. It seems that Mr Carlsson was disappointed at the way his team could not bring the Northvolt factories up to speed. In keeping with modern times, Mr Carlsson found a perfect solution - AI! Let’s just replace those pesky workers with AI and robots, so that Mr Carlsson and Mr. Khullar could run a factory very efficiently and in peace! Managing people is so tedious! Same as all investors, I, for sure, adore Mr Carlsson’s genius. I just hope that in his genial spark, he won’t be using Northvolt’s data as an input to train his AI. After all, as most AI scientists say, garbage in, garbage out. I would doubt, though, that Mr Carlsson could find a willing battery OEM to share with him the necessary data. Jokes aside, I would imagine that after a blunder as big as Northvolt, Carlsson would be persona non grata in the investment world. But it seems that I am fundamentally missing something here. It just doesn’t make sense to me that investors continue to invest not only in failed founders (that’s ok, failures happen), but in those who blew it out of all proportion, like Adam Neuman or Carlsson. If that were to happen in Russia, I would assume that Mr. Carlsson pocketed the money, shared it (kicked back) with the fund managers, and let the venture die. Nice and clean way of separating investors from their money. Then he would go and repeat it with another venture, with all the fund managers cheering him up as a founder who tried, but failed, and deserves to try again, as he now has "experience". But of course, such a thing could only happen in criminal states like Russia, it is simply implausible for it to happen in Sweden. It must be Mr. Carlsson’s vast experience that allows him to rise again and again. The most visible demonstration of this, as you might have noted, is how deftly Mr. Carlsson rode the wave of battery hype, and how timely the “batteries” theme is now replaced with the currently hot AI theme. Hype, raise, burn, repeat. Could someone please recommend outlandishly expensive online courses or Bali retreats, where I could learn the wisdom of raising billions, burning them, and raising again? Because I am definitely missing something here.

Can you build a solar/storage company with no outside capital?
 Turns out, you can. If your customers pay upfront. Earlier this week, I spoke with a few startups in B2C solar and storage.
 Some sell high-end solutions to the well-off.
 Some team up with real estate developers. 
Others build SaaS platforms connecting customers, installers, and hardware suppliers. None of this is my core territory.
 But I’ve been learning fast, and one insight stood out. Every one of these ventures faces the same issue: working capital.
 And the most common answer? Advance payments.
 One founder shared that 80–100% of the order value comes in before delivery. Think about that.
 No dilution.
 No debt.
 Just customer cash to get started. Of course, this only works when you’re small and your customers are willing to wait.
 Turns out that many are. Two, even three months. But if you want to scale—really scale—you’ll need to start offering credit.
And that’s a different game entirely. Still, if you’re building in B2C solar and storage, maybe start with this question:
Can you get your first hundred customers to pay upfront? That might be all the capital you need. How do you work your working capital?

🔋 The European battery industry has taken some serious hits over the past 12 months: Freyr pulled the plug, ACC stalled, Northvolt collapsed, and CustomCells is the latest to go under. What’s common in all these cases? High capital burn, delayed scale-up, and a gap in *deep manufacturing know-how*—especially in electrode production. Many respected voices—Christopher Chico, Michael Sura, Christoph Siara—have said it plainly: if Europe wants a battery industry, it has to work with those who’ve already mastered it. That means partnering with Korea, Japan, and yes, China. But let’s be honest—how many of those top-tier Asian battery players are really ready to share their IP or know-how? If you’re at Smarter-E Europe this week and you’re serious about tackling electrode and cell production challenges, there’s one booth you should not miss: 📍 Booth C3.640 at the ChungBuk Technopark stand. That’s where you’ll find JR Energy Solution—a Korean foundry that offers manufacturing-as-a-service for electrodes and pouch cells. 👷🏻‍♂️ What’s new this year? — 20+ chemistries produced (including LFP, NCM, Si, and solid-state trials) — A new pouch cell line, with sample modules/packs — Open contract manufacturing slots for H2 2025 — And hands-on training for manufacturing personnel at their 500 MWh facility in Eumsong, Korea. This is what Europe needs—real partnerships with those who’ve done it before, and are willing to roll up their sleeves and help. Let’s build smarter, faster, and stronger. Together. 🛠️ Let me know if you’d like an intro.

Everyone wants to decarbonize industry. Not many talk about what happens after the pilot works. Scaling hardware isn’t linear. It’s exponential — in pain. Case in point: Via Separations Their FOAK gamble? Scaling their membrane-based heat separator 100x at a commercial paper mill in Alberta, Canada. * The science? ✅ Solid * The pilot? ✅ Promising * The scale-up? ❌ A swamp of unexpected engineering nightmares: 🔧 Labour costs exploded — getting a good welder on site during snowstorms = great for the schedule, brutal for the budget 🧳 The team wore every hat imaginable and spent weeks on-site ❄️ Winter storms delayed shift changes and logistics 📦 Equipment had to be ordered on-demand, not upfront — shaving 18–24 weeks off lead times to keep the project alive This wasn’t a “tech readiness” issue. It was ops readiness, supply chain readiness, and above all, mindset readiness. FOAK means debugging the real world. Despite it all, the team got it done. Was it painful? Yes. Was it worth it? Absolutely. If you're on the FOAK frontlines, building tech for hard-to-abate sectors, let’s trade war stories. I’m building a living archive of real-life scale-up pitfalls and tested FOAK frameworks. The original interview of Shreya Dave, CEO of Via Separations, can be found here .

If you’re building your first-of-a-kind (FOAK) project, good advice is hard to find. Most online sources on climate FOAKs are either too fragmented or come with a strong VC spin—lots of talk about financing, little about building. That’s why this new guide from Climate Drift  is worth your time: 👉 The FOAK Field Guide It’s one of the most comprehensive pieces I’ve seen in a while—broad in scope, packed with case studies, practical links, and sharp observations from people who’ve actually done it. A great primer for anyone moving from pilot to demo, or from demo to FOAK. And if you’re hungry for more insights like this, I keep a regularly updated directory of articles, case studies, and field notes on FOAKs and cleantech scale-up: 👉 FOAK & Scale-Up Blog Directory More practitioner voices. Less hype. Picture credits: Climate Drift

I'm deep into research on large-scale solar heating vs. heat pumps – and the findings are fascinating. At the household level, heat pumps dominate. They’re fast to install, scalable, and increasingly efficient. But when you move into industrial heat and district heating at a serious scale, the picture flips. Solar thermal projects start pulling ahead. At scale, solar heat beats heat pumps on operating costs, emissions, and resilience. Once built, the sun’s energy is basically free. And in regions with high solar irradiation and large, steady heat demand, the economics are hard to ignore. The trade-offs? Solar thermal takes longer to build, needs a serious land area, and is location-dependent. It’s not a plug-and-play solution everywhere. But if we are serious about decarbonizing industrial and district heat, especially in Europe, we need to think beyond just heat pumps. Scaling solar heat is slow... but if you want cheap, clean heat over decades, it's probably worth the patience. Curious to hear from others: where do you see solar heating fitting into the clean heat transition? #climatetech #decarbonization #solarthermal #heating #heatpumps

Everywhere I turn — LinkedIn, podcasts, even my dumb drip coffee machine — someone’s screaming about AI. The latest gem? Eric Schmidt telling Congress that AI could consume 99% of all energy. Anyone who's ever tried getting a grid connection permit would probably fall off their chair laughing. Everyone needs to take a deep breath. We’re not in an AI revolution. We’re in an AI hysteria — whipped up to trigger FOMO, pump stock prices, and funnel cash into AI startups that may or may not do anything useful. After some breathing exercises, here’s what’s left: - AI won't build your house. - AI won't generate your electricity. - AI won't grow your food or sew your clothes. - AI won’t fix your grid, your water system, or your climate. Sure, AI will have great uses. I use ChatGPT every day for research, brainstorming, and editing. It’s a fantastic tool. But it’s not going to save the world. It’s not going to break it either. And no, it’s definitely not going to consume 99% of global energy. Meanwhile, every minute spent obsessing over "AGI taking over" is a minute not spent dealing with real problems: climate change, wars, pandemics, inequality. So take a deep breath. Fire up your LLM. And go figure out a solution to something real.

Airbus just quietly moved the launch of its hydrogen-powered plane from 2035 to 2045. In corporate speak, that’s not a delay—it’s a polite way of pulling the plug without admitting failure. After €1.7B spent, even Airbus is starting to realize that hydrogen aviation may be a science project, not a business. I called this a year ago, when Universal Hydrogen filed for bankruptcy after burning through $100M. That was a 20-minute or $100M crash course in hydrogen physics. Airbus just paid for the platinum edition. Hydrogen has its place—but it's not up in the air. Low energy density, bulky storage, safety concerns, and no existing infrastructure… The list goes on. Popcorn's out. I'm waiting for the next hydrogen aviation moonshot to quietly disappear into the clouds (ZeroAvia). In the meantime, here’s my write-up on the Universal Hydrogen saga if you missed it: 👉 https://www.askerov.pro/post/universal-hydrogen-a-20-min-or-100m-lesson #climatetech #aviation #hydrogen #scaleup #cleantech #FOAK

Selling to corporate clients is hard. Selling to utilities? That’s next-level. If you’ve ever tried to get your hardware or software into the operations of a water, power, or grid operator, you’ll know what I mean. LineVision, a U.S.-based startup, managed to do what many couldn’t: land a paying utility customer. And in doing so, they illustrated something every cleantech startup needs to internalize early—just because your product is a no-brainer technically doesn’t mean it’s an easy sell commercially. LineVision developed a solution that, on paper, looked like a silver bullet for overloaded electricity grids. Their dynamic line rating system—composed of non-contact sensors and smart analytics—could boost power line capacity by up to 40%. A simple install, real-time wind and weather monitoring, and a software interface that tells grid operators when lines can safely carry more load. Sounds perfect, right? Well, not quite. To understand why, you have to understand how utilities work. And for that, you need to forget almost everything you learned from selling to tech companies. --- Why Utilities Don’t Innovate (Until They Really, Really Have To) Utility companies are regulated entities. Their revenues are largely determined by government-set tariffs. These are based on costs, not outcomes. If they save money using your technology, it might mean lower future revenue, not higher margins. That’s a big disincentive right out of the gate. Add to that a strong internal culture of caution. Most utilities put service continuity above everything else. So the internal motto is: "If it isn’t broke, don’t fix it.” Engineers are rewarded for keeping systems running, not trying new things. That’s why most innovation pitches go nowhere. The only exceptions? When the pain becomes impossible to ignore. --- LineVision’s Real Opportunity: The Grid Is Getting Squeezed LineVision spotted that inflection point. Load growth from data centers and renewables was stressing U.S. grids, and regulators were starting to demand answers. New transmission lines take years to permit and build. So, utilities needed a Plan B. LineVision offered a faster way to unlock hidden capacity from existing infrastructure. But even then, it wasn’t a slam dunk. Their breakthrough came with Duquesne Light Company (DLC) in Pennsylvania, one of the few U.S. utilities willing to act early. And even then, it was only possible because LineVision understood how to sell to utilities, not just pitch to them. --- How They Did It: From Innovation Team to “Pilot Hell” The entry point was DLC’s innovation and strategy team. That’s where most utility partnerships begin—not in procurement or operations. The innovation team’s job is to anticipate future system needs. If your pitch solves a big enough problem, you’ll get passed to the technical team. That’s when the real test begins. LineVision pitched a live data stream from sensors to the cloud, which would feed real-time analytics to utility control rooms. Sounds smart? Not to utility engineers. The moment they mentioned “cloud-based architecture,” the room went cold. DLC’s cybersecurity protocols made cloud streaming a non-starter. Like many utilities, their control rooms are air-gapped from the internet. Not out of paranoia, but because if your grid gets hacked, bad things happen fast. So LineVision went back to the drawing board. They rebuilt their system to run offline. No cloud streaming. Just weather inputs and pre-trained models, with data from sensors used for later validation, not real-time operations. That pivot—not their original tech—is what got them the deal. --- Lessons for Climate Tech Startups 1. Utilities don’t reward efficiency. They reward reliability. Focus your pitch on operational resilience, not just savings. 2. Start with the technical case, not the business case. Until the engineering team signs off, nothing else matters. 3. Be ready to redesign. Security, compliance, and compatibility often matter more than your core innovation. 4. Timing is everything. Pain from external pressures (demand spikes, regulatory changes, infrastructure bottlenecks) creates the windows where innovation can get in. 5. Pilot Hell is real. Surviving it requires humility, persistence, and a willingness to learn from the customer’s side. --- Final Thoughts We’re not going to decarbonize the grid just by making better tech. We have to embed that tech into organizations that are, by design, skeptical and slow to change. That means understanding their incentives, their fears, and their constraints—then working within them. LineVision’s story is a masterclass in how to do just that. If you’re building for the grid—or for any other conservative, regulated infrastructure sector—don’t just ask: Is my solution technically sound? Ask: *Will a utility engineer be willing to bet their job on it?* If not, you’re not done yet. --- ✉️ Got a grid tech or cleantech scale-up challenge? Reach out. I’ve sat on both sides of the table and helped companies navigate from pitch to pilot to deployment. Let’s make your solution real. This article was written on the basis of The Green Blueprint Podcast, Episode “LineVision’s quest to win over utilities” https://www.latitudemedia.com/news/linevisions-years-long-effort-to-operationalize-new-technology-at-utilities/

I spent yesterday in my favorite mode: deep in the weeds with startups trying to scale. One minute I was analyzing a 900 MW offshore wind project with floating turbines that haven’t even been tested commercially. The next, I was helping a lithium scale-up line up potential off-takers for a low-carbon hydroxide plant still in pre-construction. Then I reviewed a predictive maintenance pitch, gave feedback on grid-scale battery storage in deep waters, and even fielded questions about a Middle East robotic crane for building super-tall lattice towers (!). This is the behind-the-scenes work of scaling climate tech. It’s not glamorous. It's rarely about pitching to VCs or tweeting about impact. It’s long calls, blunt feedback, untangling supply chains, asking the unsexy-but-crucial questions: Who’s going to buy this? Will it survive commissioning? Can it be manufactured and deployed without breaking the bank? And every now and then, connecting a founder with someone who might just help unlock their next step. Scaling climate solutions means turning great lab ideas into grid-connected, customer-ready, shovel-in-the-ground reality. If you’re somewhere along that journey, reach out. Happy to challenge, support, and help move things forward. Let’s scale the hard stuff!

Let’s be clear: blaming AI for increased electricity demand is like blaming your toaster for climate change. A kilowatt-hour is a kilowatt-hour, whether it powers ChatGPT or a hospital. Yes, AI is driving up electricity demand. But so do electric vehicles. So do growing economies. Different forecasts put AI data center energy consumption at 945-1500 TWh by 2030. During the same period, Africa's electricity demand is expected to grow by 600-700 TWh. So, Africa alone will likely consume the same volume of energy as all data centers in the world. And it's more likely that African demand and supply will hit those targets, while with AI: 🤖It is not clear whether there will be so many use cases for it (there are not many now), 💰Whether AI companies will make any money (they are not making any now), ⚡️And whether data centers will need that much energy (Deepseek case shows that maybe they won’t). The real issue isn’t who uses the energy. It’s how we build the supply. If we’re serious about scaling energy to meet global demand, we know what works: ✅ Solar ✅ Wind ✅ Batteries ✅ Maybe some natural gas for balance And what doesn’t? ❌ SMRs (still not real) ❌ Hydrogen as fuel (looks good on slides, not in practice) ❌ Concentrated solar (20 years late and still expensive) Nuclear? Sure — if someone, anyone, figures out how to build it on time and on budget. So let’s stop freaking out about AI energy use. We don’t have an AI problem. We have a supply problem. And we definitely have a hype cycle problem.

The EU wants its own battery industry. From lithium mining to recycling. From cells to EVs. 100% sustainable, 100% European. Sounds great on paper. But what does the EU actually do to make this dream happen? Not much. Sure, there’s Horizon Europe, the Innovation Fund, and the Recovery and Resilience Facility — all throwing serious money around. But none of it is laser-focused on batteries. Funds are spread thin across dozens of competing green tech bets. EV subsidies? Some EU countries have them. But they go to any EV — whether it's made in Slovakia or Shanghai. That might sound good in a free-market economics textbook. Competition = efficiency, right? But in practice? EU producers are fighting a war on two fronts: → higher energy prices → higher labour costs Two things they can’t easily control. Even if they go full sci-fi: cover every roof in Chinese solar panels and fully automate production with robots — they just trigger other problems: higher unemployment and even higher electricity prices as energy demand falls. This is not how you build a strategic industry. You want a local battery and EV supply chain? Time to go old-school: sticks and carrots. → Carrot: huge subsidies for EV buyers. Big enough that everyone wants to ditch their diesel burner. → Stick: subsidies only for EVs meeting local content requirements. Start simple — what Europe already has: Car bodies, interiors, electronics, battery packs. Then level up every 4-5 years: → Cells → Electrodes & active materials → Recycled & mined metals Keep this rolling for a decade — and you'll have a real industry, real jobs, real resilience. No shortcuts. No magic new tech. Just industrial policy done right.