Leveraging Big Data in lithium-ion battery asset management can reduce safety risks, save money and extend battery life, but all Big Data comes with challenges.

In the current era, the digital revolution has brought with it a powerful tool that is transforming the way companies operate and make decisions: Big Data. NCPOWER, a leading manufacturer of sustainable and customised lithium batteries for electromobility with headquarters in Murcia, has this technology for the optimisation of battery performance. 

NCPOWER's latest Big Data Center upgrade goes beyond simple data collection. It integrates artificial intelligence and machine learning to take decision-making to the next level. 

The role of Big Data in battery efficiency

Big Data is not simply a large amount of information; it is the ability to analyse, understand and utilise data on a monumental scale. For NCPOWER, this means that every aspect of its production, from battery design to the supply chain, is supported by advanced data analytics.

One way to understand this is to look at what the famous 5 Vs of Big Data mean: volume, variety, velocity, value and veracity. Each of these elements are essential and highlight the challenges in this area.

Let us look in more detail at the 5 V's mentioned above:

1. Volume

In the context of Big Data, "Volume" is about the magnitude of data, from Terabytes to Pettabytes. In the battery universe, battery management systems (BMS) generate data, and while a single BMS produces a manageable amount, the accumulation of historical data from multiple subsystems can reach Terabyte scales.

For example, a domestic PV storage system might have 1-4 modules (5kWh to 15kWh) sending data, while a large-scale battery storage unit might have thousands of modules and range from 50kWh to 500MWh sending data continuously. This increase in size and complexity can result in large volumes of data, presenting challenges for efficient storage and proper information management.

2. Variety

Having a large volume of data is challenging enough, but having variety within that data increases the complexity.

The data generated by batteries varies according to the application: a home storage system may provide information on current, voltage and solar energy, while an electric bus provides data on speed, power, voltage and current.

To address this complexity, it is vital to consider the variety of data from the initial stages of designing a data pipeline. Implementing concepts such as the 'data lake', a non-relational storage method, offers flexibility by supporting different data formats. This approach is essential to anticipate and manage the diversity of data that may emerge as new types of battery systems evolve and are introduced.

3. Speed

In the field of battery data processing, "Speed" is crucial to perform real-time or near real-time analysis. In battery systems, especially with more computationally intensive safety algorithms, the ability to compute quickly is essential to prevent critical failures.

Historically, the battery management system handled this speed internally, but now, with the evolution of algorithms and the need for faster analysis, cloud computing becomes essential. This allows speed to be maintained in the embedded BMS algorithms, avoiding delays in results and improving security.

4. Value

Having abundant data is only the first step; the real wealth comes from analysing this data effectively and being able to deliver value. This is where you unlock the potential to:

• Reducing security risks and related costs
• Information on key decisions in the supply chain
• Collateral monitoring
• Battery life extension

By analysing information continuously, unplanned downtime and other risks are avoided. In addition, it improves profitability and also benefits the environment.

In short, value is generated not only by having data, but by leveraging it to make informed, strategic decisions that strengthen the profitability and sustainability of battery assets:

We get clarity on the operational performance of batteries when compared to the manufacturer's specifications. 

• We analyse the expected value and whether there is an adequate return on assets.
• We see whether electrics meet the warranty conditions for end customers.
• In addition, we evaluate which battery suppliers perform best in their operating conditions and charging routines.

Continuous safety monitoring is also carried out to prevent critical failures and reduce operational risk. We use advanced ageing models based on machine learning to predict battery degradation with certainty, enabling the creation of robust business plans. 

5. Truthfulness

Data veracity can become challenging due to the wide variety and massive volumes of data. A voltage considered an outlier in one system may be completely normal in another.

To address this, it is crucial to add context to the Battery Management System (BMS) data through metadata describing the battery system. This metadata provides essential boundary conditions for future accuracy checks, ensuring that the data is interpreted correctly.

Understanding that battery data is embedded in the Big Data paradigm is the first step. Acting accordingly, by developing data channels that handle volume, variety and velocity, is critical to ensure accuracy.

Conclusion

In short, NCPower's latest Big Data Center upgrade represents not just a technological evolution, but a comprehensive strategy to address Big Data challenges. The company doesn't just collect data. It uses it effectively to make informed decisions, drive efficiency and lead the lithium-ion battery industry into a sustainable and profitable future.