Problem Statement: Industrial BESS energy management systems typically rely on simplified aging assumptions, treating battery degradation as proportional to energy throughput and neglecting operating-condition dependence. Such simplifications lead to aggressive charging and discharging strategies that reduce battery lifetime and distort the true economic value of storage, particularly under variable tariffs, temperatures, and industrial load profiles.

Approach: The proposed approach integrates physics-aware battery degradation models with AI/ML-based forecasting tools within an industrial EMS framework. Historical load demand, energy price signals, and contextual operational data are used to train data-driven models for short-term demand and dispatch prediction. These forecasts are combined with physics-informed degradation mechanisms in the EMS optimisation layer, enabling adaptive, health-aware charging and discharging decisions that balance economic objectives with long-term battery reliability.

Expected Outcome: The proposed EMS achieves lower lifecycle energy cost, reduced degradation rates, and demonstrably improved alignment between short-term operational decisions and long-term asset health.