CES 2026: The Consumer “Everything” Show – Battery Edition

Battery Electric Highlights & Technical Deep Dive 

CES used to stand for “Consumer Electronics Show.” This year, someone jokingly called it the “Consumer Everything Show” — and honestly, that feels accurate. Who would expect a fusion reactor company at CES?

Three years ago, when I attended CES for the first time, it felt like the Consumer Electrification Show: batteries, EVs, charging, power electronics everywhere. Looking back at my 2023 experience — from John Deere’s electric excavator with Kreisel’s immersion-cooled battery to Stellantis’ RAM truck with Factorial solid-state batteries — the show was dominated by pure electrification content.

This year, electrification is still foundational — but it’s now embedded into almost everything else: AI, robotics, data centers, autonomy, mobility, infrastructure.

Okay, so let’s dive into my personal tech highlights from CES 2026 – as usual, it’s all things BATTERY 🔋🤩

🔋 Energy Storage for AI Data Centers: The New Frontier

For the first time at CES, I saw multiple companies showcasing battery energy storage systems specifically designed for AI data centers — spanning UPS and backup power, on-site energy optimization (peak shaving), and grid support.

One standout example was XING Mobility‘s BBx800 — an 800V immersion-cooled high-voltage DC backup battery designed specifically for high-power AI data centers. This NMC based system delivers approximately 1.2 MW per unit and uses full immersion cooling to maintain operating temperatures between 25°C and 27°C over extended periods. This design directly addresses thermal runaway risk and improves system stability under extreme loads.

What makes this interesting from a technical perspective:

• Architecture placement matters: Requirements differ significantly depending on whether batteries sit at the grid, UPS, or conversion layer. The BBx800’s 800V configuration (supporting both ±400V and 800V) reflects this — higher voltage means lower current at the same power, which translates to reduced thermal losses and improved charging capability.
• Immersion cooling: This isn’t new to data centers, where it is already used extensively for chip-level cooling in high-density AI compute. It’s also not new to batteries in high-performance applications — think back to Kreisel’s immersion-cooled battery system in the John Deere excavator shown at CES 2023, among other examples. What’s interesting now is the transfer of these thermal management concepts from automotive and off-highway applications into data center energy infrastructure, highlighting how solutions developed in one sector are increasingly being adapted to solve challenges in another.
• The “AI data center” problem: U.S. electricity demand is projected to increase dramatically by 2030, largely driven by AI workloads. Data centers are shifting from passive energy consumers to active grid participants — co-investing in infrastructure, enabling load flexibility, and deploying on-site generation and storage. The topic came up repeatedly last year, and the Battery Show 2025 panel discussion was a particularly strong reference point (for more details read our article on Electrification Academy.)

Beyond Lithium: Technology-Agnostic Approaches

Stryten Energy showcased lead, lithium, and vanadium redox flow battery technologies at CES 2026, emphasizing a technology-agnostic approach. Their vanadium flow batteries offer safe, scalable, long-duration energy storage for data centers demanding 5-24+ hour resilience — something lithium-ion struggles with at this scale.

This diversity in chemistry reflects a maturing market. Not every application needs the highest energy density. Sometimes what you need is duration, safety, and the ability to cycle thousands of times without degradation.

🧪 Solid-State Batteries: Shown, Not Just Promised

Schaeffler: Pressure-Controlled Solid State Battery Systems in Robotics

Meeting the Schaeffler team and seeing their pressure-controlled solid-state battery system integrated into robotic applications was a personal highlight for me. I’ve been talking for years about the mechanical pressure requirement in many lithium-ion–based all-solid-state battery concepts — and it remains one of my core concerns for adoption in large scale battery applications. 

These cells require significant mechanical pressure to operate properly (around 5 MPa in Schaeffler’s case) and, at the same time, exhibit pronounced breathing behavior, with thickness increasing substantially from the uncharged to the charged state. What made Schaeffler’s booth stand out was their candor: they openly discussed a ~30% volume expansion during cycling and the mechanical integration complexity that comes with it.

That level of transparency is rare in “our space”. They showcased a solid-state battery module integrated into a humanoid robot, as well as a smaller, dog-like robotic platform moving around the booth – all including their active mechanical pressure control system.They also had a video demonstrating a cell penetration test that did not lead to thermal runaway. It’s important to note that the larger module on display was a concept, not populated with fully representative production cells.

Technical note: Volume expansion in certain solid-state battery chemistries is not a secondary effect — it directly affects mechanical stability, interfacial contact integrity between layers, and long-term cycle life. Any viable system must solve this at the pack and system level, not just at the cell.

ProLogium: Solid State Cell Manufacturing Transparency with FEV and RIMAC

ProLogium and FEV showcased a next-generation EV battery module built on ProLogium’s superfluidized all-inorganic solid-state lithium ceramic battery technology. The module targets approximately 1,000 km driving range and has already advanced to the sample car-ready stage.

What impressed me:

• Real manufacturing data: ProLogium’s Gigafactory in Taoyuan, Taiwan, has shipped more than 600,000 battery cells — this isn’t a lab concept anymore.
• System-level thinking: FEV contributes expertise in battery system design, thermal management, and BMS control strategy integration. This combined approach gives automakers flexibility: either maximize range (~1,000 km) or reduce vehicle weight by approximately 300 kg for the same range.
• Fast charging potential: With appropriate system design, the module has the potential to deliver 60-80% state-of-charge in 4-6 minutes without compromising safety.
• Ionic conductivity: ProLogium’s system achieves 57 mS/cm at room temperature — approximately 5× that of conventional liquid electrolytes.

The collaboration between ProLogium (cells), FEV (system integration), and RIMAC (vehicle platform) demonstrates how solid-state is moving from material science to complete system engineering.

Donut Lab: Bold Claims on a new type of Solid State Battery

This topic has already received a lot of attention, so instead of repeating headlines, I want to share my direct experience from CES.

Together with industry veteran Paul Charles, I had the opportunity to speak with several people from Donut Lab: Ville Piippo, their CTO and Craig Williams, their VP of Business Development (ex Rivian). 

What was shown at CES (and what wasn’t)

At the booth, Donut Lab showed models of a battery cell and a battery module — as far as I understood these were not the actual functional products. They also displayed the Verge TS Pro motorcycle, which is their first announced partner/customer application.

According to Donut Lab, shipments for the Verge TS Pro are planned to start in April 2026, but in limited volume. One of the executives mentioned on the order of ~100 motorcycles initially.

What Donut Lab said about the battery

Below are the key claims as they were communicated to us — with important context.

• Performance claims (cell-level unless stated otherwise)
  • Energy density: ~400 Wh/kg
    → Roughly double today’s typical lithium-ion EV cells.
    → They explicitly referred to this as a cell-level number.
    → They also said pack-level energy density depends on integration, including thermal management — which, according to them, is sometimes required, sometimes not, depending on the application.
  • Lifetime: 100,000 cycles
    → Explicitly stated to be MODELED, not yet field-validated as this would require 20 years of testing
  • Vehicle range: up to 600 km for the large battery version of the Verge TS Pro
  • Charging time depends on application
    • Claimed 5-minute charging with appropriate cooling (≈ 10C charge rate)
    • Important note they made themselves: the Verge TS Pro does not currently have active heating or cooling, so in that configuration charging would be closer to ~10 minutes
  • Operating temperature range: –30 °C to +100 °C with ~99% efficiency

• Additional technical characteristics they mentioned – These are the points that made me pause and ask: What kind of system is this really? A battery? A supercapacitor?
  • No self-discharge!
  • The CTO explicitly stated that the cells show no swelling and no volume change during cycling, and that no mechanical pressure or “breathing” is required — unlike many conventional solid-state concepts that rely on sustained stack compression.
  • At the same time, he mentioned that there is still some “weirdness at the pack level” that needs to be considered during integration. He did not go into detail, and it remained unclear whether this referred primarily to mechanical integration, thermal behavior, or electrical system effects when many cells are combined.
  • Voltage window matches typical NMC lithium-ion, enabling easier integration with existing BMS architectures
  • No rare-earth materials
  • Any shape and form possible, with variable cell voltage
  • They explained that they started with a pouch cell on purpose, to mimic a conventional NMC lithium-ion cell and simplify integration into existing applications.
  • They also claimed that in the long term, a BMS might not be required, though today they clearly design for compatibility.

• Manufacturing & scaling
  • They said they currently operate a facility in southern Finland with a 1 GWh annual run rate. The exact location was intentionally not disclosed.
  • According to them, the company was formed through a strategic partnership with Nordic Nano, with Donut Lab taking a controlling stake to secure proprietary IP and manufacturing know-how.
  • They claimed inbound interest from over 2,000 companies, and emphasized that they are selective, prioritizing customers who are actually going to market.
  • They did not disclose details on chemistry or manufacturing methods, but they explicitly stated: No rare-earth materials, Locally available materials, Not traditional roll-to-roll processing
  • They are targeting a 10 GWh by 2027, potentially more, depending on investment
  • Market strategy: Initial deliveries to Verge Motorcycles (hundreds of units), followed by COBA Power electric trailers (600+ kWh packs — 30× larger than the motorcycle application), and a skateboard platform collaboration using their in-wheel motors. They’re prioritizing high-margin applications first (drone defense, UAV, marine, aerospace) where safety is paramount and customers will pay 10× per unit compared to commodity lithium iron phosphate.

My take on Donut Lab — excitement, but also caution

This is not comparable to a traditional lithium-ion battery — and likely not to what most people think of as “solid-state” either.

You’ve probably seen the online speculation linking Donut Lab to Nordic Nano’s amorphous titanium-dioxide TiO₂ nanostructures, and to the work of Dr. Bela Bhuskute, Chief Scientist at Nordic Nano Group. My dear friend  Lacy from Miss Go Electric laid out one of the most thorough (speculative) analyses I’ve seen, describing a possible anode-free, sodium-based solid-state architecture with strong capacitor-like characteristics, potentially enabled by nano printing-based manufacturing techniques (SCREEN Printing).

To be clear:
This connection is still speculative. Neither Donut Lab nor Nordic Nano has released peer-reviewed data, detailed process descriptions, or independent validation results.

The most accurate statement today is this:

The Donut Lab / Nordic Nano stack appears fundamentally different from conventional lithium-ion systems and may share traits with advanced nanostructured or capacitor-like architectures — but without independent testing and formal technical publications, the true operating mechanism remains unverified.

What I like — and what still worries me — coexist very closely here.

• On the positive side, Donut Lab has real customers, not just slides, and they are clearly putting applications first.
• The CTO came across as genuine and technically grounded, and the team is deliberately starting small and focused, rather than claiming immediate mass adoption.
• I’ll admit it: I’m excited — but carefully so.
• What keeps me cautious is equally clear. There is no public cell data or third party validated cell data yet, there is no real aging data yet, as the headline 100,000-cycle lifetime is still modeled, not field-validated.
• Manufacturing yield remains a major unknown.
• And finally, there are pack-level integration risks: the CTO himself hinted at some “pack-level weirdness,” without going into detail. That could relate to current distribution, localized hotspots, or transient system behavior — exactly the kinds of issues that don’t show up in cell-level specs, but often determine whether a technology succeeds or fails AT SCALE.

⚡ Battery Charging: Less Hype, More Substance

From Chemistry-Only to Charging Strategy

Automotive fast-charging infrastructure dominated CES 2024. This year it’s wireless charging focused on consumer electronics. While automotive applications remain niche (remember Stellantis’ wireless charging robot for the RAM truck in 2023?), consumer devices are driving real innovation in inductive charging efficiency and power transfer.

Instead of chemistry-only narratives, more attention is finally going to how batteries are charged. Do we see a conversation shifting from “What’s the cathode and anode?” to “How do we optimize the charging profile in real time based on degradation behavior?” 😅

Partnership with Iontra: Adaptive Charging

I was grateful to partner with Iontra this year and dive deeper into advanced battery charging technologies. Our conversations centered on adaptive charging — where charging profiles are optimized per cell and application, in real time, based on degradation behavior (read more in this article.)

Key insights:

1. Charging strategy is just as critical as chemistry when it comes to performance, lifetime, and safety — especially in cold weather.
2. We are not just talking about WHEN to charge in terms of at which State-of-Charge, or at which Temperature – but HOW to charge? What is the exact charging profile?
3. Find the optimum adaptive charging methodology can increase lifetime by 2-3 times!
4. Optimized charging on “cheaper cells”: One insight that really stayed with me is that optimized charging routines have a particularly large impact on cells that are not manufactured to the highest possible standard.

This raised interesting questions for me around how charging control and manufacturing quality interact. If adaptive charging can compensate for manufacturing variability, what does that mean for the entire cell production process? Could we tolerate slightly higher variability if the BMS or charger is intelligent enough to handle it? (read more in this article.)

Huge thanks to Daniel Higgs and Madeleine Stump for great discussions and product launches with MAGFAST and Salom.

🧠 Battery Management: Becoming Even More Intelligent

Battery management systems are evolving from relatively simple state-of-charge estimation toward real-time interpretation and predictive decision-making. This is where consumer electronics, automotive, and silicon truly intersect.

Over the past years, I’ve been collaborating closely with Texas Instruments, and CES has consistently been one of the most valuable checkpoints to review what’s actually moving the needle in BMS technology. If you haven’t seen it yet, there’s a great video and article from earlier CES editions that’s still worth revisiting.

Texas Instruments: Wireless BMS with In-Vehicle EIS

At CES 2026, Texas Instruments showcased wireless BMS technology with in-vehicle Electrochemical Impedance Spectroscopy (EIS) — enabled by increasingly capable on-chip signal processing.

Why this matters: EIS has traditionally been a lab technique for battery characterization. Bringing it on-vehicle, in real time, means you can monitor battery health at a much deeper level — tracking impedance changes that indicate degradation before capacity fade becomes obvious. Doing this wirelessly removes wiring complexity in the pack, reducing weight and improving flexibility in module design.

AI-Driven Approaches: Electra and Others

Companies like Electra are pushing AI-driven battery management  — moving beyond physics-based models to data-driven approaches that can learn from large fleets and adapt to individual battery behavior. It was great to reconnect with their CEO Fabrizio Martini and the team (for details check out our podcast).

Reality check: AI in BMS is promising, but it’s not magic. You still need good data, robust validation, and fail-safe mechanisms. The best approaches combine physics-based models (we know how batteries degrade) with machine learning (to capture real-world variability and edge cases) and AI. 

🤖 Robotics That Feel… Normal

From Futuristic to Expected

Robotics was EVERYWHERE at CES. After the first 5-10 demonstrations, the novelty wore off. But Hyundai’s booth did something that stood out to me personally: it made autonomy and robotics feel very practical in my industry.

From autonomous charging and logistics concepts to robotic vehicle movement systems — robots stopped feeling futuristic and started feeling expected.

After a while, I found myself asking: “When wouldn’t you automate this?” instead of “Why would you automate this?”

That shift in perspective — from skepticism to assumption — feels significant.

💡 The Caterpillar Insight: Imagining What Comes After

I’ll end with a thought from the Caterpillar keynote. They showed a drawing from around 1930: an engineer imagining a future where a customer could describe a machine problem “through television,” a Caterpillar engineer could diagnose it remotely, identify the right spare part, and have it delivered — almost immediately.

At the time, people probably thought this engineer was insane.

But honestly, walking through CES 2026, that pattern felt different. It felt surprisingly easy to imagine how the near future works — robots in everyday environments, autonomous systems coordinating quietly in the background, intelligence embedded into things we already use.

What’s harder now is not imagining that future — but imagining what comes after it.

We’re at an inflection point. The technology is integrating. Quietly. Incrementally. And very seriously.

Closing Thoughts

As always, CES is as much about people as it is about technology. I’m grateful for the conversations, debates, and unexpected encounters — and I’m still processing everything I saw.

A few themes keep circulating in my mind:

1. Energy storage is diversifying — lithium isn’t the only answer, and AI data centers are driving demand for duration and safety over pure energy density.
2. Solid-state is transitioning from lab science to system engineering. The hardware exists. The constraints are real. The manufacturing challenges are acknowledged and being solved.
3. Charging and BMS are becoming first-class design considerations — not afterthoughts. How you charge and manage a battery is as important as what chemistry you choose.
4. Robotics and autonomy are normalizing — faster than I expected.

If you were at CES 2026 and saw something I missed — or if you have thoughts on any of these topics — I’d love to hear from you.

We will be back next year!

—Veronika