In this thesis, we investigate new multi-echelon vehicle routing problems for logistics operations using unmanned autonomous vehicles. This can provide immediate tangible outcomes, especially in high-demand areas that are otherwise difficult or costly to serve. This type of problem differs from the commonly used multi-echelon supply chain management systems in that here there exist no intermediate facilities that consolidate/separate products for delivery; instead all decisions are made on a per-vehicle basis. We describe here how we can obtain the necessary parameters (data collection) to evaluate the performance of such multi-echelon systems. We also provide three mathematical formulations based on different assumptions and case scenarios. We then study the differences between the three models in practice, as far as routing cost and duration of operations are concerned. We finally show that there are savings to be had by properly employing unmanned vehicles for logistics operations.