LeydenJar: Accelerating R&D and production automation with smart data integration

A data-hungry startup

LeydenJar is developing an anode that produces a new generation of batteries. They are in an intensive R&D phase where they are iteratively improving their technology. Data is essential for their innovation process: they collect data about machine speed, product quality, speed/quality ratio and component durability. This came with challenges. LeydenJar grew rapidly, but many processes had not yet been professionalized. With a limited budget, they had to make difficult choices between different investments.

The consequences were palpable: their BI specialists spent 80% of their time collecting data manually. As a result, there were few hours left to create reports, insights and AI models. All data was stored in Excel files, which made the situation error-prone, unclear and unsuitable for AI applications.

Hidden treasures in scattered data

LeydenJar collected a lot of valuable data, but did so in isolated systems. For example, they measured the “cycle life” of batteries (the number of charge-discharge cycles up to 80% capacity), collected performance from anode production machines, and performed five different quality tests on produced anodes.

All these datasets were viewed and analysed separately, making it virtually impossible to find correlations. In addition, much knowledge about R&D tests was lost due to distributed storage and the lack of documented data interpretation. This led to situations where similar R&D tests were carried out multiple times by different departments.

We decided to show the potential for improved data collection and explain what they could achieve with it. One of Leydenjar's core values is: “Question everything — answer with data.” They realized that they had much more potential in their hands. Together, we joined forces to lay a solid foundation: the LeydenJar Data Warehouse project.

From vision to concrete architecture

LeydenJar was good at expressing what they wanted to achieve. But they still had no clear idea of exactly what data they were collecting manually, what data they wanted to collect in the future, and what applications they would use it for.

We started with a high-level assessment of current and future data needs. We then drew up a longlist and shortlist of data platform providers and developed a reference architecture. By shortlisting this architecture, we were able to give well-founded advice on the best tech stack.

The choice fell on Microsoft Azure. After that, we created a Solution Design and phased the implementation in several steps so that LeydenJar could scale up according to their needs. We implemented the technique, set up the ETL process based on the Medaillon framework and added the first data sources.

Accelerating innovation and time to market

The automation of data collection provides LeydenJar with enormous efficiency gains. The data is now in a Data Lake, so it can also be used for AI applications. Their R&D process is much more efficient and faster, which speeds up time to market. This means they can grow faster from a startup that needs funding to a company that sells products and makes a profit.

We transferred all the knowledge to LeydenJar's Data Engineer, who, with our help, made Azure his own. He is now working independently to add new data sources. Their Data Scientist can now fully focus on AI and ML without wasting time collecting and transforming data.

Standardised data collection offers LeydenJar important advantages:

• Centralized storage that acts as a Single Point of Truth
• Documented datasets with clear reading guides and interpretation
• Improved data quality through better control of data input
• Automation options by reusing existing data
• Standardised and reproducible data enrichment
• Correlable datasets through data standardization
• Advanced Machine Learning opportunities to gain new insights

Through this transformation, LeydenJar has not only improved their current processes, but also laid the foundation for future innovation and growth in the competitive battery production market.

Blis Digital has proven to be an indispensable partner for LeydenJar by guiding us from setting up requirements to the successful realization of our Data Warehouse. Their speed and flexibility were crucial for us as a start-up and scale-up. The project was a success within budget and scope, with short lines of communication and close cooperation between everyone involved. Thanks to Blis Digital, we now have a robust foundation for LeydenJar to make a difference in battery technology, and the positive reactions within our company speak volumes about the impact of their approach.
Freddy van der Garde, Information Technology System Manager, LeydenJar

‍