AVL Focus - Issue 2024

attery cells are extremely sensitive, requiring extensive,

time-consuming tests during development and

production to ensure safety and performance. In

addition, various cell types, test chambers, temperatures,

and test scenarios must be considered. Many test labs

still work manually, leading to errors and delays. There are

tLree eJJective ways to coYnteract %Ytomated testfield

management, simulation-driven reduction of testing time, and

AI-based data analysis. These strategies increase traceability

and transparency, reduce costs, and shorten time to market.

1. Digitalization and Automation in the Lab

Using digital test orders and automated test procedures

reduces manual tasks, avoids errors, and improves the

quality of results. A comprehensive digital order processing

system with automated planning of the entire battery lab

assigns up to 1000 units under test to the correct test

chambers. Performance monitoring and KPI-driven control

of testbeds also reduces energy consumption. The AVL Lab

Management™ software solution seamlessly integrates

hardware and software, thus optimizing utilization, creating

transparency and traceability.

2. Simulation-Supported Test Optimization

Battery tests are very time-consuming due to slow

electrochemical processes. Our algorithms, developed over

years of collaboration with pilot customers and our own

propulsion development, analyze test data and optimize test

Tougher global competition for battery-electric vehicles requires

a shorter time-to-market for new models. Therefore, as a core

component, batteries must be developed and tested faster while

maintaining higher quality.

TrocedYres. -n Fattery cell aging tests, YT to Ż oJ time can

be saved, as not every test needs to be completed to the end

and can be supplemented by simulation. AI models learn the

dynamic behavior of batteries – which varies with temperature

and state of charge – and thus eliminate unnecessary test

cycles. Real-time simulation on the battery testbed enables

validation and optimization at system level under realistic

driving conditions in the lab (‘road-to-lab’). This reduces the

risk of late, costly changes just before start of production.

3. AI-Driven Evaluation of Test Results

Accurate prediction of the state of charge and the lifetime of

batteries requires in-depth understanding of electrochemical

processes. Fleet data is used to train data-driven models

based on neural networks to predict battery aging. Combined

with the data from cell tests, this can improve the quality of

battery models and further reduce development time. This

requires huge volumes of data to be processed from various

development areas and test environments. Our leading-edge

data analytics platform pools this data and thus serves as a

central source of knowledge.

%:0 Tlaces great emTLasis on JYtYreTroofing, reƽected in oYr

portfolio. Our solutions in the areas of virtual testing, data

analytics and testfield management are scalaFle and ƽe\iFle,

allowing them to adapt to technological changes and pave the

way to a safe and sustainable future.

Strategies for More

Efficient Battery Testing

HOME OF INNOVATION

E-MOBILITY

20 | 21