Autonomous mobile robots are one of the new and innovative ways to improve operation in industries such as warehouses, logistic companies, agricultural businesses, healthcare institutions, and a lot more. They are known for their operational improvement, safety, efficiency, and speed, automating several functionalities so they can be performed with little or no human intervention.These advantages can only be realized, however, if the degree of autonomy suffices for the task at hand. Whilst many degrees of autonomy exist (e.g., functional autonomy), in this thesis we are primarily concerned with an aspect of non-functional autonomy: energy. One of the important features of an autonomous robot system is the capability to charge autonomously with little to no human intervention. This thesis examines the energy distribution problem on multi-robot warehouse systems. We model warehouse systems where robots charge autonomously, pausing their workload when needed, to charge at a station. Depending on specific execution, it is possible that robots fully deplete their energy before arriving at a charging station, decreasing total work achieved, and becoming an obstacle on the warehouse floor. We then introduce the concept of energy sharing, where robots are capable of charging one another, essentially becoming mobile charging stations.In this context, the problem of energy distribution becomes a problem of multi-agent collaboration.We analyze the impact of our solution on a multi-agent simulation, showing that energy sharing for autonomous mobile robots in warehouse systems reduces the total number of depleted robots and contributes to increasing the amount of work performed.