Route planning for cooperative air-ground robots with fuel constraints

Abstract

Limited payload capacity on small unmanned aerial vehicles (UAVs) results in restricted flight time. In order to increase the operational range of UAVs, recent research has focused on the use of mobile ground charging stations. The cooperative route planning for both aerial and ground vehicles (GVs) is strongly coupled due to fuel constraints of the UAV, terrain constraints of the GV and the speed differential of the two vehicles. This problem is, in general, an NP-hard combinatorial optimization problem. Existing polynomialtime solution approaches make a trade-off in solution quality for large-scale scenarios and generate solutions with large relative gaps (up to 50 %) from known lower bounds. In this work, we employ a hybrid metaheuristic known as Construct, Merge, Solve & Adapt (CMSA) in order to develop a scalable and computationally efficient solution approach. We discuss results for large scale scenarios and provide a comparative analysis with the current state-of-the-art.

Energy-Aware Drone Path Finding with a Fixed-Trajectory Ground Vehicle

“…They propose treating all drone waypoints as a single TSP, then breaking up the route into sub-tours. This work was further expanded in [8,9] where they propose a mixed-integer linear programming (MILP) solution. A similar problem is found in [10], where they first plan ground vehicle routes along a road network then plan drone routes using Conflict-Based Search.…”

Unreliable Sensors for Reliable Efficient Robots

“…This possibly leads to r 1 .x + r 2 .x 2 = r 1 .x, so r 1 stops moving and the defeat condition for Convergence is wrongly activated. We test this by setting r 1 .y = r 2 .y = 0, picking r 1 .x at random in [0, 1] and picking r 2 .x at random in [2,3] so that r 1 .x < r 2 .x.…”

“…The overwhelming majority of the mobile robots research has focused on proving, under a given set of conditions, whether there exists a counter example to a given problem. On the other hand, the practical efficiency of a given algorithm (with respect to real-world criteria such as fuel consumption) was rarely studied by the distributed computing community, albeit commanded by the robotics community [2,45].…”

“…In fact, an overwhelming majority of the research on mobile robotic swarms has focused on proving, under a given set of conditions, whether there exists a counter example to a given solution proposal for a problem. On the other hand, the practical efficiency of a given algorithm (with respect to real-world criteria such as fuel consumption) was rarely studied by the Distributed Computing community, albeit being of paramount importance to the Robotics community [2,45]. Fuel-constrained robots have been considered in the discrete graph context, for both exploration [24] and distributed package delivery [11], but, to our knowledge, no study considered the two-dimensional Euclidean space model that was promoted by Suzuki and Yamashita [41].…”

A Self-Adaptive Variant of CMSA: Application to the Minimum Positive Influence Dominating Set Problem

“…CMSA has been successfully applied to a number of combinatorial optimisation problems. Some of the latest applications include the one to the maximum happy vertices problem [17], to route planning for cooperative air-ground robots [18], to refuelling and maintenance planning of nuclear power plants [19], and to the prioritised pairwise test data generation problem in software product lines [20].…”