Can we bound the time it takes to complete the task based on inter-site distances, energy requirements, and number of robots?
Is there an optimal robot replacement policy for a given problem?
Can we tweak the roadmaps in some way to improve the replacement efficiency? For example, what happens if we split undirected edges into two directed pieces to create 'road lanes'?
Can we make algorithms that are robust to robot/battery failures?
How do the properties of the resources affect things? What happens if we vary the simulated weight, number of robots required to move them, add multiple types requiring different drop-off sites, include things like 'hazardous' materials that require special paths, etc.?