Energy-Efficient Motion Planning for Multi-Modal Hybrid Locomotion

Abstract

Hybrid locomotion, which combines multiple modalities of locomotion within a single robot, can enable robots to carry out complex tasks in diverse environments. This paper presents a novel method of combining graph search and trajectory optimization for planning multi-modal locomotion trajectories. We also introduce methods that allow the method to work tractably in higher dimensional state spaces. Through the examples of a hybrid double-integrator, amphibious robot, and the flying-driving drone, we show that our planner tractably gives full-state trajectories that are probabilistically optimal and dynamically feasible.

Towards reconfigurable and flexible multirotors

“…In this context, there are many indoor mapping techniques proposed: a safe interval path planning and integer linear program with a 3D indoor map for a hybrid locomotor (Araki et al 2017), a graph-based planning using a 2D map (Sharif et al 2019). Some algorithms account for system dynamic constraints in their hybrid vehicles motionplanning algorithm using approximate dynamic programming (Terry Suh et al 2020). However, most multi-modal planning algorithms are demonstrated for aerial-ground vehicles and not for aerial-aqua vehicles.…”

Energy-Efficient Autonomous Navigation of Solar-Powered UAVs for Surveillance of Mobile Ground Targets in Urban Environments

“…Air-ground robots (AGR), which are known for their outstanding mobility and long endurance, have been gaining significant interest lately and show great potential for applications in search and rescue tasks [1]- [3]. Existing works [4]- [6] have demonstrated success in fast air-ground hybrid path planning, particularly in simple and unobstructed scenarios. However, AGR navigating complex environments (e.g., forests or buildings) with occluded and unknown areas faces a dilemma since obstacles in these areas significantly affect the results of path planning, i.e., high collision probability and suboptimal energy consumption (in Fig.…”

Towards autonomous multi-modal mobility morphobot (M4) robot: traversability estimation and 3D path planning

“…To map the environment and path planning, Sharif et al [39] utilize a 3d uniform gridbased method and a Araki et al [40] also utilize a 3d uniform grid-based method but it has fixed a flight height for the robot. On the other hand, sampling methods are employed in [41,42] to generate a graph-based representation of space for robots that can fly and do wheeled motion. The current state-the-art for multimodal robots path planner is Eric Sihite et al [10] which uses a 3d MM-prm method to discretize the environment and sample it.…”

“…To compute costs on the graph's edges, most articles consider factors such as power consumption [39,40]. Suh et al [41] propose an optimization method using machine learning based on a reduced physical dynamics model. The optimization function accounts for constant power drain, battery voltage, and motor torque constant.…”

“…Additionally, in the work by [43], path costs are computed by formulating risk in various scenarios.also in the work for 3d For graph searching, Dijkstra's Algorithm [44] or its improved version, A ⋆ , is commonly implemented in most studies. A post-processing step involving a smoothing algorithm [41] is often applied to optimize the path.…”