The global shift towards carbon neutrality is transforming the automotive industry by boosting electric vehicle (EV) adoption. These EVs rely on lithium-ion batteries (LiB) due to their high energy density and long cycle life. However, their capacity degrades over time, making the monitoring of state of health (SoH) crucial. Current SoH estimation methods include physics-based, model-based, and data-driven approaches, with a combination of these, often referred to as hybrid models, offering potential improvements. Much of the current research on SoH estimation overlooks real-world applications and relies on laboratory data that does not accurately reflect field conditions. In reality, the quality of field datasets is far from ideal which makes it incompatible with most state-of-the-art data-driven techniques. This paper, therefore, focuses on highlighting the limitations of laboratory data and reviewing the practical challenges faced when using field battery data for SoH estimation. A method that uses specific constraints is, subsequently, presented to estimate baseline SoH values using only partial charge data. Additionally, current data-driven techniques that use field battery data are analysed to identify their limitations and advantages.