Precise state-of-health (SOH) estimation is extremely challenging in lithium-ion based cells due to the nonlinear behavior of its operating parameters. This pertains to complex estimation and prediction characteristics of power and energy demands in hybrid electric vehicles (HEVs) and electric vehicles (EVs), particularly variations in the operating temperature. In this paper, agricultural electric power tiller application battery pack cells are utilized for the study analysis. Extended Kalman filter algorithm (EKF) analysis was carried out for the experimental cell voltage estimation. The estimated voltage applied as the input to the nonlinear autoregressive network (NARX) modeled to analyze actual versus predicted charge/discharge powers. These estimation studies were carried out for lithium-ion nickel manganese cobalt chemistry 18650 cells. Aging analysis was carried out for four study cells through electrochemical impedance spectroscopy (EIS) to measure the impedance growth with different SOC conditions under a constant temperature of 318 K. In addition, incremental capacity analysis (ICA) peak points studies were carried out with regression analysis performed to evaluate the change in capacity. Resistance characterization analysis with two different temperature conditions 308 K and 318 K was performed to evaluate the change in resistance for the four study cells.