FABBS offers batteries an energetic second life

Eight ambitious AI startups have been selected to compete in the AI Pitch Competition. The finals are on November 13, 2025, and IO+ will portray each contestant in the run-up to that event. The AI Pitch Competition is a Brabant-based contest that highlights the most innovative AI solutions, offering startups the opportunity to present their ideas, connect with industry leaders, and accelerate their growth. Today, we show what FABBS has in store for the world. Co-founder and CEO Justin van den Hurk answers our questions.

What specific AI technology is at the core of your solution?

"We use machine learning to get parameters that fully define the state of a battery cell, allowing us to get an accurate SoC, SoH, SoP, predict remaining useful life, and degradation mechanisms, among others. Currently, there are a couple of methods to define the state of a battery cell:

1. ECM (Equivalent Circuit Model) parameters: A low-fidelity model of a battery cell can be made using resistance and capacitance values to define an electric circuit. This is the current standard in battery management. The current-voltage interaction of this model is then assumed to simulate that of the battery. A problem here is that the relation is highly temperature-dependent, but this is not part of the model, so look-up tables are used for values at different temperatures. In reality, these parameters do not describe the full state of the cell; they only provide a temperature.
2. A physics-based Full Order Model (FOM): A high-fidelity model of a battery can be made by using the differential equations from conservation laws on the microscopic scale of a battery. These equations are numerically solved using a spatial pseudo-2D discretization method, which is like dividing the battery up into many small pieces and taking average values for each little piece. This gives very good results, but it is computationally expensive. Getting the parameters of these models requires cell teardown.
3. A physics-based Reduced Order Model (ROM): a high-fidelity model of a battery similar to the FOM, but with some simplifications that make it solvable.

Having these parameters allows us to better control how we load each cell, which we can adjust with our balancing circuit. These parameters can be determined using physics-based approaches, but this process is quite challenging.

Our data pipeline looks like:

• Our power converter circuit performs real-time (lab-grade) cell measurements of current-voltage interactions.
• Using those measurements to train/find parameters that best predict the measurements.
• Get data on how accurate the model's state estimates were.
• Learn which profiles to send with our power converter circuit."

How scalable is your AI solution?

"Our solution is highly scalable through a standardized data pipeline running directly on battery management systems (BMS). The model adapts to various battery chemistries and system configurations using our parameter learning framework.

Key challenges include ensuring consistent performance across different chemistries, managing limited compute power on embedded systems, and obtaining initial training data. But once our system is deployed on batteries, it collects more data that can be used to train improved models and gain more insight into battery cell design or optimization of battery use."

How does your startup address potential ethical concerns related to bias, fairness, or transparency in AI decision-making?

"We ensure fairness and transparency by keeping our models chemistry-agnostic and documenting the data origin of every dataset used. All data collection and processing are done with explicit customer consent and stored securely."

In what ways do you believe your AI solution can positively impact society?

"Our AI solution makes a societal impact in three ways:

1. With our innovation, we enable second-life applications, significantly reducing the global reliance on rare earth elements and lowering waste.
2. Improving performance, like increasing capacity by 20%, allowing for 10-20% faster charging, and reducing the overall total cost of ownership of batteries. This promotes the adoption of batteries and speeds up electrification.
3. By doubling overall battery life."

Tell us more about your entrepreneurial journey

"Our biggest hurdle up until now was finding the funding to execute our idea and grow our team. We have not fully overcome this yet, as this is a constant challenge, but we have secured some funding through grants, and we are working on obtaining some debt financing."

How are you preparing for the increasing regulatory frameworks around AI, such as GDPR, AI Act, or other data privacy laws?

"We are fully aligned with GDPR principles and preparing for compliance with the upcoming EU AI Act. All data is stored securely by FABBS, with clear documentation and consent processes. Our AI models include traceable versioning and data documentation to ensure accountability. Compliance enhances our credibility and encourages transparent innovation rather than limiting it."

Future vision: What is your long-term vision for your AI solution?

"Over the next 5–10 years, we aim to deploy our systems in a wide range of batteries, actively pushing those batteries to their absolute physical limit. All the while, collecting large sets of data, more sophisticated than previously possible, as input for better cell and battery design.

Our technology will drive electrification in aviation, maritime, and heavy-duty sectors, and foster a circular economy by ensuring second-life batteries are as reliable as first-life ones. We see our company at the forefront of physics-informed AI for sustainable energy systems."

What opportunities does your location give you? What is still missing in the ecosystem you are part of?

"Being located in Brabant, we benefit from the high-tech Brainport ecosystem and the strong academic network of TU/e. Collaboration with partners like NXP, LionVolt, VEDS, AME, VDL, and DAF gives us access to cutting-edge hardware, testing, and industrial expertise.

What’s still missing is close access to large battery cell manufacturers, which are mostly located in Asia, but more keep coming to Europe and the Brainport specifically as well."

The AI Pitch Competition: Why will you win this contest?

"We will win because our solution uniquely combines physics-based modeling, machine learning, and active hardware control to deliver measurable improvements in battery performance, lifespan, and sustainability.nIt is a tangible, high-impact innovation grounded in Europe’s strongest deep-tech ecosystem, with a strong path to commercialization and societal benefit."