toplogo
Sign In

Coverage Planning for Energy-Constrained UAV and UGV with Mobile Recharging


Core Concepts
Efficient coverage path planning for energy-constrained UAV and UGV teams with mobile recharging.
Abstract
The paper presents an approach for coverage path planning for a team of an energy-constrained Unmanned Aerial Vehicle (UAV) and an Unmanned Ground Vehicle (UGV). The goal is to achieve complete coverage by both robots while minimizing the coverage time. The UGV can also function as a mobile recharging station, and the UAV and UGV need to rendezvous occasionally for recharging. The proposed heuristic method addresses this NP-Hard planning problem by initially determining coverage paths without considering energy constraints. Subsequently, segments of these paths are clustered, and graph matching is used to assign UAV clusters to UGV clusters for efficient recharging management. Numerical analysis on real-world applications shows that compared to a greedy approach, the proposed method reduces rendezvous overhead on average by 11.33%. A proof-of-concept demonstration is provided with a VOXL m500 drone and a Clearpath Jackal ground vehicle, showcasing the system from offline algorithm to field execution.
Stats
Our method reduces rendezvous overhead on average by 11.33% compared to a greedy approach. Up to 25% improvement in some instances was observed.
Quotes
"The proposed heuristic method addresses this NP-Hard planning problem." "Our method reduces rendezvous overhead on average by 11.33%." "A proof-of-concept demonstration is provided with a VOXL m500 drone and a Clearpath Jackal ground vehicle."

Key Insights Distilled From

by Nare Karapet... at arxiv.org 03-18-2024

https://arxiv.org/pdf/2310.07621.pdf
AG-CVG

Deeper Inquiries

How can stochastic energy consumption be addressed in the context of this coverage path planning

In the context of coverage path planning with energy-constrained UAVs and UGVs, addressing stochastic energy consumption involves considering uncertainties in the energy usage of the vehicles. To handle this, one approach is to incorporate risk-aware online decision-making methods into the algorithm. By modeling the energy consumption as a stochastic process with probabilistic distributions, such as Markov Decision Processes (MDPs) or chance-constrained optimization, the system can dynamically adapt to changing energy levels and uncertainties during operation. This allows for more robust and adaptive decision-making regarding recharging rendezvous locations based on real-time information about energy usage.

What are the potential limitations of using an offline approach like AG-CVG in real-world scenarios

Using an offline approach like AG-CVG in real-world scenarios may have some limitations that need to be considered. One potential limitation is that offline algorithms do not account for dynamic changes in the environment or vehicle conditions during operation. In real-world scenarios, factors like unexpected obstacles, varying weather conditions affecting battery performance, or deviations from planned trajectories due to external disturbances could impact the effectiveness of precomputed paths. Additionally, offline approaches may not be able to optimize paths in response to real-time data feedback or adjust strategies based on evolving mission requirements.

How can risk-aware online decision-making methods be integrated into the clustering and matching process of AG-CVG

Integrating risk-aware online decision-making methods into the clustering and matching process of AG-CVG can enhance its adaptability and efficiency in handling stochastic energy consumption scenarios. By incorporating risk assessment metrics into cluster formation and graph matching stages, the algorithm can prioritize clusters based on their associated risks related to uncertain energy levels or environmental conditions. This integration enables AG-CVG to dynamically adjust rendezvous plans based on current risk evaluations while optimizing coverage paths for both UAVs and UGVs under uncertainty constraints. Risk-aware techniques ensure that decisions are made considering both optimal path planning objectives and potential risks associated with stochastic variables present during operations.
0