Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
The battery company pivoting to AI is SES AI, formerly Solid Energy. It is shifting emphasis away from trying to compete in high-volume electric-vehicle cell manufacturing and toward Molecular Universe, an AI platform for battery-material research, alongside battery materials, drones and energy storage. SES is not leaving batteries: it is betting that its data and chemistry expertise may be more valuable when sold through software, services and materials than through mass-produced EV cells.
That is both a technology strategy and a response to hard manufacturing economics. The platform could help researchers find and test candidate materials faster, but a promising molecule is not a qualified, affordable battery. Whether SES’s pivot succeeds depends on evidence of customer adoption, repeatable lab results and materials that can be manufactured at scale.
What SES is changing—and what it is not
SES AI is a Massachusetts-based company founded out of MIT research by CEO Qichao Hu. It began as Solid Energy in 2012, developing advanced lithium-metal batteries, and later pursued silicon-anode technology and electric-vehicle applications. The company has worked with automakers including GM, Hyundai and Honda; those historical relationships should not be read as proof of current production contracts.
In 2026, SES began putting greater emphasis on Molecular Universe, its battery-focused AI-for-science platform. SES describes a broader business that also includes drones, energy storage (ESS), materials and battery-related products. The clearest description is a change in where it hopes to earn value—not a clean exit from battery hardware.
#1 Best Overall
![Arduino® UNO™ Q 4GB [ABX00173]- Hybrid Board, Qualcomm Dragonwing QRB2210 microprocessor (MPU) & STM32U585 Microcontroller(MCU), AI Vision, Voice, IoT, Robotics, Linux Debian OS, Wi-Fi 5, USB-C](https://m.media-amazon.com/images/I/51cl4xygwGL._SL160_.jpg)
- Dual-Brain Hybrid Power: Combines the Qualcomm Dragonwing QRB2210 MPU (Quad-core Arm Cortex-A53 @ 2.0 GHz CPU, Adreno GPU, AI acceleration) and the real-time, low-power STM32U585 MCU for advanced applications like object recognition, voice commands, and motion detection.
- AI & Linux Capabilities: Unlocks AI-powered vision and sound solutions; runs Linux Debian OS for coding in Python and supports the Arduino ecosystem with libraries and Sketches; quick start with Arduino App Lab.
- Advanced Features: Equipped with 4 GB LPDDR4 RAM, 32 GB eMMC built-in storage, ideal for single-board computer (SBC) mode, running multiple simultaneous high-level processes, more complex AI or ML models, extensive logs. Dual-band Wi-Fi 5 (2.4/5 GHz), Bluetooth 5.1, and high-speed headers for vision, audio, and display peripherals.
- Seamless Expansion & Connectivity: Features the classic UNO form factor for shields compatibility, an 8x13 LED matrix, and a Qwiic connector for easy expansion with Modulino nodes; power and connect via the USB-C connector.
- Intended Use & Development: The perfect platform for prototyping robotics or IoT projects, empowering innovators with a unified development experience to mix Arduino Sketches, Python scripts, and containerized AI models in a single interface.
Current picture: SES presents Molecular Universe as a platform for battery research and development, with cloud and on-premises options, and advertises software subscriptions, services and materials supply. Its site lists newer platform offerings, including MU-3.0 and MU-StarSeeker. These are company descriptions and announcements; they do not establish revenue scale, customer count or widespread commercial deployment. SES does not display public pricing on the product page.
How SES got here
SES’s early work addressed batteries for oil-and-gas exploration sensors operating above 120°C. It developed solid-polymer lithium-metal technology, then pursued the much larger EV market. The company built pilot capacity in Massachusetts and Shanghai and, during the 2021 battery investment boom, sought partnerships with major automakers.
In 2022, SES announced a move toward a silicon anode, presenting it as a potentially easier route to manufacturing than lithium metal. That was a strategic direction, not proof that the chemistry had achieved commercial success. The company’s current AI emphasis builds on that history: it is seeking to turn years of battery experiments and domain knowledge into a discovery and development platform.
Free tools Windows power users keep installed
One-click scans. No signup required.
Why building EV cells is such a difficult business
Battery manufacturing presents two different tests: making a cell work and making a business work. A promising prototype does not automatically become a dependable factory product. Cell makers must produce consistent quality, meet stringent safety and performance requirements, deliver reliably and support products over long periods. Automakers need confidence not only in a cell’s performance but also in supply, quality control and warranty risk.
Factories and production processes demand substantial capital. A startup also needs enough volume to spread fixed costs, while competing with established manufacturers that have mature plants, supplier relationships, purchasing power and hard-won process expertise. A technically better cell can still lose if it costs too much or proves difficult to manufacture consistently.
Rank #2
- Dual-Brain Hybrid Power: Combines the Qualcomm Dragonwing QRB2210 MPU (Quad-core Arm Cortex-A53 @ 2.0 GHz CPU, Adreno GPU, AI acceleration) and the real-time, low-power STM32U585 MCU for advanced applications like object recognition, voice commands, and motion detection.
- AI & Linux Capabilities: Unlocks AI-powered vision and sound solutions; runs Linux Debian OS for coding in Python and supports the Arduino ecosystem with libraries and Sketches; quick start with Arduino App Lab.
- Advanced Features: Equipped with 2 GB LPDDR4 RAM, 16 GB eMMC built-in storage, ideal to develop in PC-connected mode, running the OS, Python scripts, and basic network services (SSH) without a demanding GUI or heavy multitasking; great for lightweight AI and memory-optimized TinyML applications, needing local storage for basic OS and core libraries. Dual-band Wi-Fi 5 (2.4/5 GHz), Bluetooth 5.1, and high-speed headers for vision, audio, and display peripherals.
- Seamless Expansion & Connectivity: Features the classic UNO form factor for shields compatibility, an 8x13 LED matrix, and a Qwiic connector for easy expansion with Modulino nodes; power and connect via the USB-C connector.
- Intended Use & Development: The perfect platform for prototyping robotics or IoT projects, empowering innovators with a unified development experience to mix Arduino Sketches, Python scripts, and containerized AI models in a single interface.
Demand adds another uncertainty. EV adoption and investment have not advanced uniformly, and U.S. policy support has become less predictable. The end of U.S. consumer EV tax credits in late 2025 is one factor cited in reporting on the weaker demand outlook, not a single explanation for it. For a startup planning around large, long-term production commitments, uncertainty about customer demand can make an already capital-intensive strategy harder to finance.
That pressure is broader than SES. Western battery startups face formidable competition on cost and scale from Asian manufacturers, and some U.S. companies have failed or changed direction. SES CEO Qichao Hu has described the environment for Western manufacturing as unsustainable for many companies. That is management’s assessment, not a complete accounting of why every startup struggles.
Recommended Free Tools
What “pivoting to AI” means at SES
Molecular Universe is not a general-purpose chatbot. SES presents it as an AI-for-science system intended to support battery research from an initial question through material discovery, laboratory testing and possible supply. Its advertised workflow combines software and physical experimentation:
- Ask: An agentic language model draws on public and private battery-domain knowledge to help frame a research question.
- Search: A battery-focused database helps identify relevant molecules and materials.
- Formulate and test: The platform is designed to connect candidate selection with automated dry and wet laboratory work, including synthesis and screening.
- Design and simulate: SES describes property-prediction models and simulation workflows, including density functional theory (DFT) and molecular dynamics-related tools.
- Predict cell performance: Machine-learning models are intended to help predict cell performance and quality.
- Validate and supply: Testing, pilot production and potential materials supply are meant to connect research results to manufacturing.
In shorthand, the ambition is: research question → candidate molecule → formulation → lab test → cell evaluation → manufacturable material. AI may help prioritize which candidates to investigate; it cannot make the need for physical testing, cell fabrication and qualification disappear. SES advertises cloud and on-premises deployments, optional compute infrastructure and services as ways for organizations to access the platform. The public product descriptions do not independently prove how well each element performs in customer settings.
A concrete example: additives for silicon anodes
Silicon can store more lithium than graphite, but it expands substantially as a battery charges and discharges. That expansion can damage the electrode and undermine performance over repeated cycles. Fluoroethylene carbonate (FEC) is commonly used as an electrolyte additive to help form a protective layer on silicon. Reporting on SES’s work notes a concern that FEC can degrade at high temperatures and generate gases that harm battery life.
Rank #3
- Single core ARM Cortex-A7 32-bit core, integrated with NEON and FPU
- Built in Micro's self-developed 4th generation NPU, with high computational accuracy and support for mixed quantization of int4, int8, and int16. Among them, int8 has a computing power of 0.5 TOPS and int4 has a computing power of up to 1.0 TOPS
- Built in self-developed 3rd generation ISP3.2, supports 4 million pixels, and supports various image enhancement and correction algorithms such as HDR, WDR, and multi-level denoisin
- It has powerful encoding performance, supports intelligent encoding, adapts to save bit rates according to the scene, and saves more than 50% of the bit rate compared to conventional CBR mode, making the captured images high-definition, smaller in size, and doubling the storage space
- The design with built-in RISC-V MCU supports low-power fast startup, 250ms fast capture, and simultaneous loading of AI model library, enabling facial recognition to be completed within 1 second
SES says Molecular Universe identified a compound intended to provide similar benefits without that gas-generation problem. The company has also said the platform identified six new electrolyte materials. These are company-reported results, not independently established commercial outcomes. A candidate is not yet proof of improved long-term cycle life, safety, cost or manufacturability. Nor does a material that works in one lab setup necessarily perform the same way in a different cell format or at production scale.
Why software and materials could be a better bet
SES’s business case is that it can reuse battery expertise and data across customers and applications rather than bear the full cost of building a mass-production cell business. Potential offerings include:
- Software access: Subscription or enterprise platform access for research teams.
- On-premises deployment: An option for organizations that need tighter control over sensitive data or computing environments.
- Research services: Paid discovery, simulation, testing or laboratory support.
- Materials licensing or supply: A route to earning value from a promising compound without manufacturing every finished cell itself.
- Selected battery products: Cells or systems for applications where performance matters more than competing solely on EV-scale cost.
Software and licensing can potentially reach multiple manufacturers without requiring SES to build a gigawatt-hour-scale factory. But “less capital-intensive” does not mean asset-free: AI-for-science still requires scientists, data, laboratories, compute, cell testing and pilot production. If customers need extensive hands-on research support, the business could look more like a specialized services company than a high-margin software vendor.
The addressable problems may also extend beyond passenger EVs. Drones, robotics, aerospace and specialty energy storage can value low weight, energy density, power or domestic sourcing enough to tolerate higher cell prices or lower volumes. Those niches could offer a practical route for battery hardware while the platform serves a wider set of R&D customers. They are possibilities, not guaranteed markets for SES.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.SES’s proposed advantage—and the hard questions
SES argues that its edge is not simply an AI model. The company points to battery chemistry expertise, proprietary experimental data, years of making and testing cells, a battery-focused database and automated lab capabilities. The combination matters: a model can propose candidates, but experiments can show whether a candidate can actually be synthesized and used in a cell.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Rank #4
- 【POWERFUL ESP32‑S3 CONTROLLER】Built‑in Xtensa 32‑bit LX7 dual‑core processor, 512KB SRAM, 8MB PSRAM, 16MB Flash for stable AI voice computing and multitask processing.
- 【Preloaded Dual AI Platforms】Comespre-installed with complete Deepseek and OpenAI voice dialogue projects.Experience intelligent voice interaction instantly. (Note: OpenAI functionality requires your own API key.)
- 【STABLE WIRELESS & CLEAR AUDIO】Integrated 2.4GHz Wi‑Fi + Bluetooth 5 (LE); dedicated audio decoding module for natural, responsive voice interaction.
- 【USER‑FRIENDLY VISUAL & PLUG‑AND‑PLAY】2” TFT‑SPI color screen shows real‑time chat; modular design, no extra wiring, ready to use after setup.
- 【FULL LEARNING SUPPORT】45 programmable GPIOs, rich interfaces, online web tutorials, free technical support for beginners & developers.
That is a plausible thesis, not a demonstrated moat. The key questions are whether SES’s data is large, clean and diverse enough to generalize across chemistries, cell formats and test conditions; whether customers will trust an outside platform with sensitive research; and whether the results can be reproduced outside SES’s own labs. Competitors can build models and automated labs, too. A head start in data is valuable only if it yields better or faster outcomes customers will pay for.
There is also a strategic question about the industry’s actual bottleneck. Faster discovery helps if suitable materials are difficult to find. But if the binding constraints are factory investment, demand, supply chains, qualification or manufacturing execution, an AI platform cannot by itself solve them. Battery investor Kara Rodby of Volta Energy Technologies has questioned whether new materials are the industry’s central obstacle amid weak investment and broader market conditions. Discovery may be useful without being the constraint that unlocks the whole sector.
How to tell whether the pivot is working
For investors and industry watchers, “AI-powered” is less informative than proof of technical and commercial progress. Useful evidence would include:
- Independent validation: Results reproduced by customers or external laboratories, with clear test conditions.
- Full-cell performance: Improvements demonstrated across relevant measures such as cycle life, charging, safety and operating temperature—not just a promising simulation or early screening result.
- Manufacturing readiness: Evidence that a material can be produced consistently with available precursors and without prohibitive cost, equipment changes or safety and environmental issues.
- Customer adoption: Named paying customers, repeat business, recurring software revenue or disclosed licensing and supply agreements—not simply access announcements or demonstrations.
- Qualification and deployment: Evidence that materials move beyond discovery into customer cell validation and production processes.
- Business durability: A clearer view of revenue from software, services, materials, drones and storage, and whether those activities reduce dependence on capital-intensive manufacturing.
Each stage matters because a molecule can look good in a simulation and fail during synthesis; a material can work in a coin cell and fail in a pouch or cylindrical cell; and a cell-level improvement can still be too costly or difficult to qualify. Battery data can also be noisy and sensitive to test protocols. Customers may require on-premises systems or data isolation, complicating efforts to build models that learn across projects.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteThe bottom line: a technology bet and a survival strategy
SES’s move is both a real technology pivot and a strategic response to the economics of battery manufacturing. It is investing in a platform that combines AI, simulation, laboratory workflows and testing, while seeking value in materials, services and selected battery businesses. At the same time, it is reducing reliance on the capital-heavy race to mass-produce EV cells.
The bet is that battery expertise and data can be monetized more flexibly than a new factory can. It is not yet evidence that AI has solved battery discovery, that SES’s materials are commercially qualified, or that the platform has become a scaled business. The decisive test is whether customers can use its tools to produce independently validated, manufacturable improvements—and whether they will pay for them.
Sources: MIT Technology Review’s March 25, 2026 report on SES; SES AI’s company site and Molecular Universe product page.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
