The increasing number of vehicles in Jakarta, Indonesia, has had a negative impact on the environment. If this trend continues, it may significantly harm public health. In response to this issue, the government has introduced mass transportation solutions, such as the Jabodebek light rail transit (LRT) system. One of the key technical challenges in operating the LRT is ensuring smooth and reliable traction motor control. This study presents a simulation of the Jabodebek LRT’s traction motor performance when traversing a hilly route with a 29 ‰ gradient. A field-oriented control (FOC) method is implemented to regulate motor speed. The train operates under a constant load, with its weight gradually increasing from the lowest to the highest point of the slope. Two tuning methods are applied to optimize the controller parameters: manual (hand-tuning) and artificial intelligence-based optimization using the Firefly algorithm and the Grey Wolf optimizer (GWO). The integral of time multiplied by absolute error (ITAE) is used as the objective function to evaluate the speed control performance. The simulation results show that the Grey Wolf optimizer delivers the best performance, achieving stable speed control despite load disturbances. The optimal proportional and integral gains obtained are Kp = 16.233861 and Ki = 0.526774, respectively.