In fact, machine learning analyses data using statistical theories and harnesses the processing power of computers to implement every statistical technique. The large amount of data we can provide means that the battery behaviour information extracted is incredibly precise and reliable. We can practically recreate a “digital twin” for each lithium battery and use it for tests and simulations, especially at the design stage, and for formulating and testing behaviour hypotheses for next-generation batteries.
What type of battery data does Flash Data Center analyse?
Flash Data Center analyses all the operating parameters of the lithium battery, specifically:
- SOH (State of Health) of the battery, to get an accurate picture of the condition of every single lithium battery out there in the market.
- Analysis of charging/discharging times with special attention paid to the minimum and maximum values reached during these activities
The benefits of Flash Data Center translate into the real optimisation of industrial applications, starting with performance
A lithium battery can be used in a wide variety of contexts and in many different ways, which all affect its longevity.
Knowing ahead of time where and how to act to extend its life cycle or improve its performance is in itself a great advantage.
Since 2012, we have produced lithium batteries for many very different contexts subject to a variety of stresses. Take, for example, the automated logistics sector, where AGVs and LGVs operate non-stop, 24 hours a day in industrial plants with temperatures ranging from -30°C to +45°C.
Thanks to remote monitoring and predictive data analysis, it is possible to:
- Understand the way in which users make use of lithium batteries
- Estimate the effective battery life on the vehicle or machinery
- Predict how the battery will behave in the future
Three important pieces of information, for us and the customer, who receives accurate planning of replacements and valuable recommendations for correcting and improving the performance of the machinery, which, of course, increases productivity.
But the benefits of remote monitoring and predictive analysis don’t end here! A typical customer that chooses Flash Battery doesn’t install just one lithium battery on its industrial machine or vehicle. We are dealing with complete systems where computer technology, IoT and artificial intelligence have become real allies in ensuring the application stays reliable over time.
Correctly sizing the application
Implementing the remote monitoring function right at the prototype stage helps, for example, to size the vehicle more accurately. Analysing the prototype data helps us understand if the application requires a lithium battery that delivers more or less energy or with specific performance-enhancing characteristics.
Getting useful information on the actual use of the application
Let’s consider that our customers are usually not the end users of the electrified vehicles. Flash Battery lithium batteries are installed on machines and vehicles that our customers sell all over the world. So, it is clear that having real-time data on their operation is a fast-track way to assess if the end user is using the application correctly (for example, if the vehicle is being exposed to repeated full discharges or out-of-range temperatures).
Let’s use a road sweeper as an example: through constant data monitoring, we can analyse how the battery is performing in that specific usage context, its charge/discharge profile, its limitations, and how the end user is employing it. With this information to hand, the producer can evaluate actual use and implement improvements to its fleets going forward, and we in turn can produce specialist batteries with increasingly better performance.
Managing advanced action planning
Lastly, Flash Data Center, with its machine learning-driven automatic data monitoring, allows the advanced planning of extraordinary maintenance work. This avoids unnecessary and expensive machine downtime and lets customers manage the end-of-life of systems independently, meaning they can sustainably plan the replacement of end-of-life lithium batteries in entire fleets and, as a result, optimise and reduce the cost of disposal and make the process of re-using components more efficient.
A system that learns from, stores and processes data for increasingly smart lithium batteries
Today, Flash Data Center can output a very accurate SOH analysis that helps us in our research into increasingly higher-performance and smarter new generation lithium batteries with ad-hoc features for the needs and use of every kind of industrial application.
Its evolution going forward will go hand in hand with the amount of data it will be provided with. We are talking about an intelligent system that is constantly learning, storing and processing. As new data arrives, the system processes and analyses the data; it is always studying and adapting to new data.
“The amount of data we are collecting compared to the past is increasing at a faster and faster rate; today, we are analysing up to 4,000 sensors per lithium battery, and the trend is growing. We are therefore well on our way to building a very sound infrastructure that can continuously process huge amounts of data, interpret trends, variations and anomalies, and reproduce realistic usage scenarios we can use to develop smarter and smarter lithium batteries.”