In this paper, an inbound material-collection system with one warehouse, multiple supplies and multiple items is
studied. A greedy randomized adaptive search procedure with VLSN for an integrated inventory-routing problem in a
deterministic setting is developed. The solutions obtained from the GRASP with VLSN are compared with the lower
bound obtained from a column generation approach. The computational results indicate that the GRASP with VLSN can
find near-optimal solutions quickly with an average error bound less than 3% for most cases. For future research, a more
efficient constructive heuristic will be developed to reduce the percent deviation from LB. Moreover, to get a strong
support to the results of this paper, GRASP with VLSN will be implemented to solve larger problem sizes of IRP and
also various types of problems.