Skip to content

Intelligent, non-destructive battery performance monitoring uses both artificial intelligence and empirical models for monitoring and verifying battery health in the short and long-term - without resorting to impractical, time-consuming and destructive testing procedures.

Batteries power a wide range of devices and systems, including phones, computers, cars, IoT devices and energy storage stations. Experience shows there is a need to understand different batteries’ performance in order to select the right battery system during the energy storage design phase. There is also a need to factor in degradation trends to estimate project investment and revenue. In addition, there is a need to estimate battery capacity to compare this with the original equipment manufacturer’s warranty declaration, and also to perform periodical check-ups to ensure assets are operated properly and safely. 

The standard practice is to do experimental testing. However, this method suffers from many drawbacks, as it is time-consuming - for example, the product qualification program (PQP) test can take several months or years to complete; destructive, for example full charge/discharge, fast charge/discharge, high/low temperature tests; and impractical in an operating situation. Lastly, testing cannot cover various scenarios due to complex operating conditions and the variability of batteries. 

“Technologies that make energy use more efficient, such as artificial intelligence (AI) and the Internet of Things (IoT), are rapidly evolving. When these technologies and batteries are well combined, the revolution in environmental and energy technologies, or the ET revolution, will occur”.

  • Akira Yishino,
  • Nobel Prize (2019),
  • Lithium-ion Battery Prototype Creator

AI powered capacity and lifespan verification 

The project aims to provide timely and accurate estimation for battery health status and lifetime prediction without additional intervention to energy storage systems or destructive laboratory experimental testing. Combining millions of hours of laboratory testing data with cutting-edge artificial intelligence technologies, this challenging problem can be addressed in two phases: 

  • Phase I: battery degradation analysis tool, providing degradation predictions for both preset and dynamic conditions. It can be used as a design assistant tool to select battery and size battery systems. For dynamic conditions, it can be used as an analysis tool to understand degradation for operational sites and estimate the remaining useful life cycles. For experimental testing side, it can speed up the testing process by using the early degradation data. 
  • Phase II: online battery health monitoring, providing a check-up of battery capacity in real-time. This type of monitoring enhances services such as comparison of warranty declaration, early warning of unusual degradation, reduce or eliminate time-consuming and intrusive capacity testing. 

The benefits can assist technicians, engineers, consultants and asset owners with battery energy storage system design, battery type selection, operating profiles analysis and comparison of battery warranties. These benefits are enabled by taking into account different batteries’ capacity degradation in different conditions; degradation in preset temperature, charging/discharging rate, averaged state of charge (SOC) and SOC window conditions; degradation analysis in dynamic conditions by providing historical operating profiles; battery performance data management; and capacity check-up in real-time.

Intelligent, non-destructive battery performance monitoring