This study is focused on the design and operation of a freight network that supports carrier collaboration where two or more carriers form an alliance and share pick-up and delivery of jobs. In carrier collaboration, it is assumed that carriers are allowed to retain some of the pickup and delivery jobs they receive from clients while releasing the rest of the jobs to a common pool. A two-stage model and framework for the application of this model in the real world are developed. The stage-1 model is for strategic planning and the stage-2 model is for operational planning. The strategic model uses demand forecasts to determine the intermodal terminal (IMT) locations and provides the pickup and delivery jobs to the operational model. The strategic model is used for long term planning, whereas the operational model is used for short term planning. The realized/actual shipping data from the operational model can provide feedback to the forecasting model to update supply and demand forecasts. The updated supply and demand forecasts are used to re-evaluate the long-term plan by opening or closing IMT locations as deemed necessary by the decision makers. The objective of the multi-period strategic model is to determine the number and location of IMTs that minimize the total relevant transportation and operational costs. The objective of the operational model is to jointly determine the optimal allocation of jobs from the common pool to the carriers and pickup/delivery routes for each truck. Numerical experiments are conducted using hypothetical networks for both models. Findings from the strategic model show regions with higher supply or demand of freight volume tend to have higher utilized IMTs and impact the total network cost most. The sensitivity analysis for budget shows that intermodal shipping share and total network cost converge at a point and the model does not add new IMTs. The alternate optimal solutions show the tradeoff between intermodal shipping share and total network cost with budget investment in opening IMTs. Findings from the operational model confirm the expectation that carrier collaboration can yield a significant reduction in the total cost of serving all pickup and delivery jobs. The cost savings from the collaboration is dependent on the spatial distribution of the nodes in the network, the network size, the distance between carrier depots, the percentage of pooled jobs, the size of the overlapping region of carriers, and the number of jobs in the overlapping region.
