In a controlled test environment, everything appears correct. Functions operate as expected, system limits behave properly, the battery charges normally, and the vehicle runs without issues. Validation tests are passed and the system is technically considered successful.
However, after months of real-world operation, the situation often begins to change. Systems that initially appear stable may start showing unexpected behaviors, and engineering teams begin to observe anomalies that were never visible during validation testing. Interestingly, these problems tend to follow very similar patterns across many EV and ESS programs.
Typical BMS issues observed in the field
- SoC drift / SoC jump: state of charge estimation gradually drifts or occasionally jumps unexpectedly.
- Incorrect SoP limits: power limits may be applied too early, too late, or not applied at all.
- False OV / UV trips: voltage protection thresholds may trigger even though the cell state does not justify it.
- Charge session drops: charging sessions may unexpectedly terminate due to BMS behavior.
- Isolation alarms and false trips.
- Sensor or measurement drift.
- Unstable fault-to-recovery behavior.
- Missing thermal runaway early detection.
- Balancing strategies becoming insufficient over time.
- Contactor and precharge control issues.
- Functional safety architecture gaps.
The typical reaction inside many engineering teams is immediate: "there must be a software bug." In reality, the issue is often not a simple software bug. The root cause is frequently a static BMS software architecture that cannot adapt to the evolving behavior of the battery over time.
The critical window: 6 to 24 months
The true behavior of a battery system often becomes visible only after months of field operation. The first 6 to 24 months are particularly important because the system experiences multiple environmental and operational stress factors at once.
During this period, the system typically encounters:
- At least one summer and one winter cycle
- Hundreds of fast charge cycles
- High temperature parking combined with high SoC storage
- Increasing cell imbalance
- Sensor drift beginning to appear
- Initial system assumptions gradually losing validity
At the same time, a silent but critical process is occurring inside the battery.
Battery aging is the silent transformation
Electrochemical aging gradually changes how battery cells behave. These changes are often invisible at system level in the early stages but become highly relevant for the BMS. Key mechanisms include:
- Growth of the SEI layer
- Increasing internal impedance
- Changes in effective DC internal resistance
- Altered polarization behavior
These processes directly affect the battery's power capability, thermal response, voltage behavior and charging characteristics.
If the BMS cannot properly track these changes and adapt its internal parameters, system decisions will gradually become less accurate. Furthermore, if cell characterization data does not sufficiently cover load, temperature, time and aging dimensions, model accuracy deteriorates significantly. If the input is wrong, the output cannot be correct.
The eMOBINO approach
At eMOBINO, we do not see a BMS as just a collection of software functions. Instead, we design it as a behavioral system that evolves over time. This requires several architectural principles:
- Advanced and adaptive BMS software architecture, capable of tracking battery behavior changes throughout aging.
- Advanced SoC, SoH and SoP estimation.
- Thermal runaway early detection.
- Model-based development, with an AUTOSAR compatible architecture and V-cycle development methodology.
- Advanced verification through MiL, SiL, HiL and fault injection validation strategies.
- ASPICE-aware development.
- ISO 26262 functional safety approach.
- Scalable BMS solutions (12V to 1500V), suitable for EV and ESS platforms.
The real question
A BMS can work on day one. It can pass validation tests. It can appear stable in the field for a while.
Does your BMS truly understand how the battery will behave as it ages?
mustafa.simsek@emobino.com
www.emobino.com
