The trailer-truck backing-up problem has long been accepted as a benchmark to test control algorithms. Its unique blend of nonlinear dynamics coupled with physical limiting conditions makes it a challenging problem. A recent control scheme used on this system took advantage of some path planning techniques. However, this scheme required re-calculation of the control, when the initial conditions changed. The control scheme introduced here will use the data generated by the path planning approach to train a neural network that will be able to create the path data from any initial condition. Since then the path information can be updated in real-time as the trailer-truck traveled, the controller is reduced to tracking the path supplied by the neural network.