Definition
SOH (State of Health) refers to the health status of a battery, indicating the degree of battery aging and remaining lifespan. In battery swapping scenarios, it is used for battery health assessment and inventory management.
Evaluation Metrics
SOH is usually characterized by capacity decay or internal resistance increase. A common definition is: SOH = Current Maximum Available Capacity / Rated Capacity × 100%. BMS estimates SOH through cycle count, capacity calibration, internal resistance measurement, open-circuit voltage curves, etc. When SOH falls below a threshold (e.g., 80%), the battery needs to be retired or downgraded for use. Scenarios such as battery swapping and leasing require high SOH accuracy, directly affecting asset value and safety risks.
Reporting by Communication Modules
Electric vehicles, battery swapping cabinets, and energy storage systems report SOH along with SOC to the cloud platform via 4G/LTE Cat.1 modules. Battery swapping operators arrange battery circulation based on SOH: high SOH batteries are prioritized for deployment, while low SOH batteries are recycled for secondary use or scrapped. EV charging stations and vehicles can use SOH to optimize charging strategies, avoiding overcharging of aged batteries. Modules ensure stable reporting of SOH data, supporting full lifecycle management of batteries.
Coordination with BMS and SOC
BMS is responsible for the collection and estimation of SOC and SOH; communication modules are responsible for data uploading. SOH is the result of long-term operation of the BMS and needs to be updated in combination with historical data and models. When estimating SOC, SOH correction must be considered; the capacity baseline of aging batteries will be adjusted with the decline of SOH to ensure the accuracy of range display and charging management.
Engineering Practice Supplement
Regarding "SOH," it is recommended to architecturally distinguish between real-time control links and non-real-time telemetry links: control and interlocks should be locally closed-loop as much as possible, while cellular networks are used for monitoring, configuration, and OTA. On the cloud platform side, observability metrics (connection duration, message latency distribution, failure codes, reconnection reasons) should drive optimization rather than just checking "connectivity." Documentation delivery should include interface specifications, error code tables, extreme operating conditions, and known limitations; test delivery should include test cases, data, and conclusions to facilitate third-party retests or audits.
The following are additional tips related to cellular IoT project delivery: it is recommended to keep screen recordings of typical scenarios, screenshots of key indicators, and summaries of masked packet captures during the acceptance phase; in the mass production phase, change records should be kept for software and hardware versions, calibration parameters, and key component batches to facilitate problem attribution, customer audits, and long-term maintenance. When payment or personal information is involved, also check whether log masking and key rotation strategies are executable.
Health Rewrites the Capacity Baseline, Not Instant Charge
State of health describes how far a battery has declined from its factory usable capacity or internal resistance. It changes slowly. A common form is current maximum usable capacity relative to rated capacity. The change in DC resistance relative to the initial value can be a supporting figure. It does not answer how much charge is left now; that is SOC. It answers how much can still be delivered after a full charge. When the module reports the management system's health together with the charge estimate, the cloud can put runtime, retirement, and second-life use on the same capacity baseline. Lead-acid, ternary lithium, and lithium iron phosphate fade along different paths and cannot share one retirement line. The chemistry differences are in IoT battery chemistry.
The value is usable only after the battery has completed the charge, discharge, or rest calibration the algorithm requires, the temperature is inside the declared window, and history was not lost in a power cut. A new cell that reports very low at first power-up is usually an uncalibrated capacity baseline or shunt, not a cell that has already aged. Reporting can be daily or per cycle; it does not have to ride on every transmission. After over-temperature, overcharge, or long storage, send an update promptly. When the figure is used to sort assets for battery swap or rental, the message needs battery identity, calibration time, and temperature, or the same pack will be recorded as two health values in two cabinets.
Converting cycle count directly into health ignores shallow cycles, high-temperature storage, and charge cutoff voltage. If a resistance measurement includes connector contact resistance, a loose terminal is treated as cell aging. If the module smooths or rounds the value itself, a slow decline becomes a step and retirement is late. A cloud that applies one fixed threshold and does not record the definition will not match asset value when one side uses capacity and the other uses resistance.
Capacity calibration is best done as a full charge and discharge at controlled current and temperature, not fitted from fragmented load. During calibration, communications can slow down so transmit heat does not change cell temperature. Nonvolatile storage of health must survive a power loss mid-write. If it is corrupt, return to unknown rather than keep using an old number that still looks normal.