State of Health (SoH)

The primary indicator of SoH is capacity fade—the irreversible loss of a battery's ability to store charge. A battery with 80% SoH can only hold 80% of its original rated capacity in ampere-hours (Ah). This degradation is driven by chemical mechanisms like lithium plating, solid electrolyte interphase (SEI) growth, and active material loss. In fleet operations, capacity fade directly reduces an agent's operational range and requires recalibration of energy-aware routing algorithms.

SoH is inversely correlated with internal resistance (DCIR). As a battery ages, its internal impedance rises due to electrolyte decomposition and electrode corrosion. This increase causes higher I²R losses during discharge, reducing usable energy and generating excess heat. A healthy lithium-ion cell might have a DCIR of 30-50 mΩ; a degraded cell can exceed 100 mΩ. Fleet orchestration platforms must account for this resistance to accurately predict voltage sag under load.

Direct SoH measurement is impossible during operation; it must be estimated through proxy techniques:

• Coulomb Counting: Integrating charge/discharge current over a full cycle to measure actual capacity.
• Electrochemical Impedance Spectroscopy (EIS): Injecting a small AC signal to measure impedance across a frequency spectrum.
• Kalman Filtering & Adaptive Models: Using recursive state estimators to fuse voltage, current, and temperature data with a battery degradation model.
• Data-Driven Methods: Training neural networks on historical cycling data to predict SoH from partial charge curves.

In heterogeneous fleet orchestration, SoH is a hard constraint for task allocation. An agent with degraded SoH has a reduced effective energy buffer and a lower maximum C-Rate capability. The Battery Constraint Solver must dynamically adjust task assignments to prevent deep discharges on weak batteries, which accelerates further degradation. Agents with low SoH may be restricted to light-duty cycles or routed to tasks near charging stations to enable opportunity charging.

The industry-standard End-of-Life (EoL) threshold for most lithium-ion batteries is 70-80% SoH. Below this point, capacity fade accelerates non-linearly, and the risk of thermal runaway increases. Remaining Useful Life (RUL) is a predictive metric derived from the SoH degradation trajectory, estimating the number of cycles or operational hours until the EoL threshold is reached. Fleet managers use RUL to schedule proactive battery replacement and avoid unplanned downtime.

State of Health (SoH) and State of Charge (SoC) are fundamentally different metrics that are often confused:

• SoH is a long-term, irreversible measure of a battery's storage capability (capacity fade).
• SoC is a short-term, reversible measure of a battery's current energy content. A battery can have 100% SoC but only 70% SoH, meaning it is fully charged but can only deliver 70% of its original runtime. Accurate fleet energy management requires tracking both metrics independently.

A predictive forecast of how many charge-discharge cycles or operational hours remain before a battery fails to meet a minimum performance threshold.

• Typically defined as the point where SoH drops below 70-80% of original capacity.
• Estimation methods: Physics-based models, data-driven regression, and hybrid approaches using Kalman filters or neural networks.
• Fleet orchestration value: RUL predictions enable proactive battery replacement scheduling, preventing mid-shift failures.

A mathematical representation of capacity fade over time, accounting for stress factors that accelerate SoH decline.

• Key degradation stressors:
  • High C-Rate charging/discharging
  • Elevated operating temperatures
  • Deep Depth of Discharge (DoD) cycles
  • Extended time at high SoC
• Model types: Empirical (cycle-counting), electrochemical (first-principles), and machine learning-based.
• Integration: Degradation models feed into energy cost functions to optimize charging schedules that minimize long-term battery wear.

A composite metric that aggregates multiple degradation indicators into a single health score, often expressed as a percentage.

• Typical inputs:
  • Capacity fade (SoH)
  • Internal resistance increase
  • Self-discharge rate
  • Coulombic efficiency decline
• Advantage over raw SoH: BHI provides a more holistic view of battery condition, capturing issues that pure capacity measurement might miss.
• Use case: Fleet dashboards use BHI for at-a-glance health monitoring across hundreds of agents.

The software interface that exposes battery telemetry to the orchestration layer, enabling real-time SoH-aware decision making.

• Data provided: SoC, SoH, cell voltages, temperatures, charge/discharge current limits.
• Commands accepted: Maximum charge rate limits, balancing instructions, safety cutoffs.
• Protocols: CAN bus, Modbus, or MQTT for IoT integration.
• Fleet integration: The BMS API is the critical data pipeline that transforms raw battery signals into actionable scheduling constraints.

Strategic planning of how and when batteries are cycled to maximize lifespan while meeting operational demands.

• Core principle: Shallow cycles (e.g., 30-70% SoC) cause less degradation than deep cycles (0-100%).
• Optimization levers:
  • Limiting maximum charge voltage
  • Avoiding full discharges
  • Reducing charge rates when thermal conditions are poor
• SoH connection: Every cycle optimization decision directly impacts the rate of SoH decline, creating a trade-off between immediate throughput and long-term asset preservation.