Today, sensors and cameras are often used to monitor the movement and behavior of pedestrians, especially where there are a huge number of visitors. The classical usage of such devices, for example in a theme park, is to identify the queue size in front of each attraction and thereby to predict the time it takes until the visit can be completed. Work has been done in the past to use statistical data that resembles the data collected by such devices to simulate the pedestrian behavior. As a result, the congestions as well as the queue sizes at different times can be predicted. This work aims in using the data obtained from the simulation to optimally schedule a list of tasks to be executed as well as to find an optimal path between each destination. As an example, one might think of a scenario where a customer enters a theme park would wish to visit as many attractions as possible in the alloted time or a large clinic where a patient has to be routed through various departments such as registration, OP, X-Ray, ward, etc. The problem involves finding the optimal sequence of the tasks and determining the fastest path between the destinations, both combined. Since the data varies over time, the problem is time dependent or dynamic. In the past, several methods have been proposed to solve dynamic shortest path algorithms and scheduling problems. However, due to the stochastic nature of the available data, it is not necessary to find the best schedule and route that takes the minimum amount of time but, it is rather important to find an optimal solution in a short time. In this paper, we study and compare different combinatorial optimization methods and heuristics that can used to determine an optimal schedule.