Many variants of SBRP have been presented during the last decades. A comprehensive description of the school bus routing problem and a relatively broad survey can be found in Park and Kim (2010). In their study, they categorized the existing works in the literature on SBRP in five categories: data preparation, bus stop selection, bus route generation, school bell time and bus scheduling. Considering the objective functions used in the SBRP, the majority of researchers aim at minimizing the number of buses used in the solution and the total distances traveled by buses. Only a few SBRP studies have considered load balancing and maximum route length as an objective function. Some applications could be found in Angel et al. (1972); Newton and…show more content… In the LAR strategy, the bus stops are determined in the first phase, and then students are assigned to those stops. Afterward, bus routes are generated. (The bus stops are selected first after which a routine process takes place). In this strategy, the allocation of students is done before the routes are created. In other words, both the bus stops selection and the student allocation to the stops are independent of the route generation, resulting in drawbacks such as an exceeding number of routes. For more details, the reader is referred to (Bodin and Berman, 1979; Dulac et al., 1980; Desrosiers et al., 1981 and…show more content… (1995) where a multi-objective formulation of the urban school bus routing problem is presented. The problem includes the minimization of several objectives such as the number of routes, the total route length, variety of students assigned per each route, the variation in route length, and the total walking distance covered by students from their homes to the bus stops. The advantage of applying the ARL strategy is that it allows an efficient load balance in allocating the students to each cluster. Moreover, the number of bus routes can be kept to the minimum level because both objectives mentioned above (minimization of the number of routes and the load balancing) are independent of both the bus stop locations and the routes generated to serve these stops. The only drawback concerns the objective of balancing the route length. In fact, due to the potential dispersion of students within a cluster, the ARL strategy cannot efficiently cope with this