Accurate estimation of lithium-ion battery State of Health (SoH) is critical for emerging applications such as reconfigurable battery systems. Although data-driven machine learning methods are promising, they often rely on costly, time-intensive aging experiments and extensive feature engineering. This work proposes a lightweight SoH-prediction framework validated on both physics-informed synthetic aging data and the NASA battery aging dataset. We evaluated Random Forest (RF) and Feedforward Neural Network (FNN) models that use only a limited number of samples from an early segment of the raw discharge voltage curve as input. Results show that RF consistently outperforms FNN across input sizes in deterministic or noise-free environments, achieving an RMSE of 0.07% SoH using just 5 voltage samples. In inherently stochastic experimental data, however, FNN can achieve an RMSE 50% lower than RF (1.28 vs. 2.87), but requires 37× more mathematical operations per inference. These findings emphasize the predictive value of the early-discharge-voltage region and demonstrate that compact, low-feature-complexity models can deliver accurate SoH estimates. Overall, the approach supports a goal of combining informed synthetic data with limited real measurements to build robust, scalable SoH predictors, reducing dependence on labor-intensive degradation testing and feature-heavy pipelines.

Grid-interactive efficient buildings (GEBs) can reduce electricity costs by coordinating loads and distributed energy resources, but repeated battery use can accelerate degradation and increase replacement cost. This study evaluates reinforcement learning (RL) control of heating, ventilation, and air conditioning (HVAC) and a battery energy storage system (BESS) in a small commercial building over 10 years with cumulative battery degradation. A high-fidelity EnergyPlus-Python co-simulation emulator was developed for a building with photovoltaic generation and BESS. HVAC-only, battery-only, and coordinated RL were compared with rule-based control (RBC), and the analysis was extended to different PV and battery sizes and time-of-use (ToU) tariffs. HVAC-only RL reduced electricity cost by 38.3% without increasing battery replacement cost and produced the lowest combined electricity and battery replacement cost. Battery-only RL achieved the largest electricity-cost reduction of 63.2%, but increased battery replacement cost by 66.7%, resulting in a combined cost higher than RBC. Coordinated control retained most of the electricity-cost reduction, at 60.8%, while reducing the combined cost by 15.3% relative to battery-only RL. Battery degradation also changed with system size because the battery operated under different SoC and cycling conditions. In addition, when the on-peak period occurred after PV generation had declined, the additional battery replacement cost became greater than the electricity-cost saving. These results show that the long-term cost benefit of smart control depends strongly on how the battery is used to achieve electricity-cost savings.

Residential buildings are emerging as candidate spaces for the adoption of smart technologies, yet their specific needs and constraints remain underexplored compared to commercial settings. This review critically examines technologies that have already been successfully implemented or show strong potential for integration into residential environments. It highlights control strategies such as rule-based control, model predictive control, and reinforcement learning and their interplay with cutting-edge tools like the Internet of Things, smart sensors and virtual sensing in home environments. This paper uniquely focuses on the challenges and opportunities specific to residential applications, providing actionable insights into overcoming these barriers, and unlike prior work that emphasizes broader sustainability goals, it prioritizes the practical implications and feasibility of deploying these technologies in homes, offering a structured framework for understanding how these innovations can meet the nuanced demands of such spaces. Our review shows sustained growth in this literature since 2016, a predominance of experimental studies complemented by simulation-based and hybrid approaches, and relatively little work leveraging open public datasets. Activity is geographically concentrated in the United States, with strong contributions from East Asia and Europe. Across the residential context, the most persistent challenges are affordability and scalability, privacy and interoperability, and the limited generalizability of case-study deployments in occupied homes. Looking forward, the most promising directions combine occupant-centric MPC/RL with IoT-enabled infrastructures, alongside standardized evaluation protocols, shared datasets, and clearer evidence on costs to support scalable adoption.