Tsukada Lab. Computer Network Group
← Publications Int'l conference

Trajectory Planning for UAV-Based Smart Farming Using Imitation-Based Triple Deep Q-Learning

Quanxi Zhou, Wencan Mao, Tomás Couso Coddou, Manabu Tsukada, Yunling Liu, Yusheng Ji

IEEE International Conference on Robotics & Automation (ICRA 2026), Vienna, Austria, Jun 1, 2026

Summary in three lines

  • The work plans UAV trajectories for weed detection, recognition and wireless sensor data collection.
  • ITDQN combines imitation of successful agents with a mediator Q-network for more stable learning.
  • Relative to double DQN, experiments reported higher weed-recognition and data-collection rates.

Figures from the paper

Figure 2: Imitation-based triple DQN
Figure 2. Imitation-based triple DQNSource: Figure 2 from Zhou et al., IEEE International Conference on Robotics & Automation (ICRA 2026) (2026) (authors’ figure from the manuscript)The diagram shows how imitation and action-value learning are combined for agricultural flight planning.

Abstract

Unmanned aerial vehicles (UAVs) have emerged as a promising auxiliary platform for smart agriculture, capable of simultaneously performing weed detection, recognition, and data collection from wireless sensors. However, trajectory planning for UAV-based smart agriculture is challenging due to the high uncertainty of the environment, partial observations, and limited battery capacity of UAVs. To address these issues, we formulate the trajectory planning problem as a Markov decision process (MDP) and leverage multi-agent reinforcement learning (MARL) to solve it. Furthermore, we propose a novel imitation-based triple deep Q-network (ITDQN) algorithm, which employs an elite imitation mechanism to reduce exploration costs and utilizes a mediator Q-network over a double deep Q-network (DDQN) to accelerate and stabilize training and improve performance. Experimental results in both simulated and real-world environments demonstrate the effectiveness of our solution. Moreover, our proposed ITDQN outperforms DDQN by 4.43% in weed recognition rate and 6.94% in data collection rate.

Video

Related research and software

Related news

How to cite

Q. Zhou, W. Mao, T. C. Coddou, M. Tsukada, Y. Liu, and Y. Ji, “Trajectory Planning for UAV-Based Smart Farming Using Imitation-Based Triple Deep Q-Learning,” in IEEE International Conference on Robotics & Automation (ICRA 2026), Vienna, Austria, 2026.

BibTeX

@inproceedings{Zhou2026b,
  title         = {Trajectory Planning for UAV-Based Smart Farming Using Imitation-Based Triple Deep Q-Learning},
  author        = {Quanxi Zhou and Wencan Mao and Tomás Couso Coddou and Manabu Tsukada and Yunling Liu and Yusheng Ji},
  url           = {https://youtu.be/D-vz9HafnUM https://arxiv.org/abs/2512.18604},
  eprint        = {2512.18604},
  archiveprefix = {arXiv},
  primaryclass  = {cs.LG},
  year          = {2026},
  date          = {2026-06-01},
  urldate       = {2026-06-01},
  booktitle     = {IEEE International Conference on Robotics & Automation (ICRA 2026)},
  address       = {Vienna, Austria},
  abstract      = {Unmanned aerial vehicles (UAVs) have emerged as a promising auxiliary platform for smart agriculture, capable of simultaneously performing weed detection, recognition, and data collection from wireless sensors. However, trajectory planning for UAV-based smart agriculture is challenging due to the high uncertainty of the environment, partial observations, and limited battery capacity of UAVs. To address these issues, we formulate the trajectory planning problem as a Markov decision process (MDP) and leverage multi-agent reinforcement learning (MARL) to solve it. Furthermore, we propose a novel imitation-based triple deep Q-network (ITDQN) algorithm, which employs an elite imitation mechanism to reduce exploration costs and utilizes a mediator Q-network over a double deep Q-network (DDQN) to accelerate and stabilize training and improve performance. Experimental results in both simulated and real-world environments demonstrate the effectiveness of our solution. Moreover, our proposed ITDQN outperforms DDQN by 4.43% in weed recognition rate and 6.94% in data collection rate.},
  keywords      = {Other},
  pubstate      = {published},
  tppubtype     = {inproceedings}
}