Autonomous systems (ASs) have design-level goals and, accordingly, strategies for achieving those goals. It requires systematic integration and guidance techniques to use their capabilities for collaborative work. Dealing with capability integration and decision guidance are some of the challenges considered in system of systems engineering (SoSE). We propose a simulation-based recommendation generation designed to influence the decision-making of ASs that are involved in a collaborative work. Towards this end, we have developed a JADE-based software module that can be used to simulate and analyze collaboration behaviors and decision patterns. The analysis results in a set of recommendations for actions that can be used to achieve collaborative tasks. In this paper, we present the application of the proposed approach to guide collaboration between ASs that participate in a high-rise building fire incident - a representative mass casualty incident response system. Based on the system simulation results, certain ASs show change in level of belonging to the collaborative work as a result of the recommended actions, which is considered as indicater to the better chance of achieving the SoS-level goal.