On SAT-Based Approaches for Multi-Agent Path Finding with the Sum-of-Costs Objective.