Sensors-based Distributed Control of Multi-Drone Systems for Agile Cooperative Aerial Manipulation

Publié le
Centre Inria de l'Université de Rennes
Unité de recherche
IRISA - UMR 6074
Description du sujet de la thèse

Over the past decade, there has been a surge in the exploration of aerial robots able to perform challenging physical interaction tasks. However, the inherent limitations in the payload capacity of individual drones have prompted researchers to explore the potential of collaborative efforts among teams of aerial robots [8]. This collaborative approach is envisioned to revolutionize various application domains, including construction, inspection, maintenance, and beyond. One of the preferred solutions to enable the aerial manipulation/transportation of objects is using cables or tethers to suspend leads to the robots (see the figure). This solution is lightweight and decouples the attitude dynamics of the aerial robots to the one of the load, which in turns increase the stability of the system.

Full pose manipulation of a cable-suspended load using multiple UAVs is a promising technique for a huge variety of future industrial applications. However, the physical interactions between UAVs, load and cables render collaborative manipulation a challenging task from both a planning and control perspective.  Existing solutions have focused on one hand, on quasi-static regimes that limit the dynamic behavior and capabilities of the system [3]. On the other hand, most solutions are centralized [1,6] or consider access to system-wide information (poses, forces, etc.) [1], which reduces the autonomy and robustness of the system and limits the applicability of these solutions to relevant real-world scenarios.

Research Objectives: The primary objective of this Ph.D. thesis is to explore sensor-based and distributed coordination strategies for multi-aerial robot systems with cable-suspended loads, facilitating collaborative object manipulation and transportation through local interactions. Distributed solutions pose particular challenges, especially when addressing communication constraints among the robots. The objective is then to consider hierarchical strategies where robots communicate at a low frequency and coordinate at a higher/planning level, subsequently executing the plan through local implicit communication based on sensor-based feedback such as vision and/or force sensing.

Scientific Challenges and Solutions: Envisioned solutions will build upon existing centralized or kinematic results [2,3] and communication-less regulation approaches [4] to propose a fully sensor-based, dynamics-based, and distributed framework for collaborative agile manipulation of cable-suspended loads. For the control side, a starting point are the existing centralized approaches based on Model Predictive Control (MPC) for single- [9] and multi-aerial robots [1,6]. Our team has undertaken preliminary work exploring the extension of [1] through a distributed MPC solution based on [9], initially at a kinematic level. Should this endeavor yield promising results, a potential trajectory involves advancing the algorithm to operate at a dynamic level. For the sensing side, the starting point will be the works on sensor-based collaborative global state estimation for multi-robot systems such as [5].

Experimental validation: The devised coordination strategies for the manipulation and transportation of cable-suspended loads will undergo thorough validation and testing using the cable-driven platform, shown in the figure, already present at Rainbow.

  1. S. Sun and A. Franchi, "Nonlinear MPC for Full-Pose Manipulation of a Cable-Suspended Load using Multiple UAVs," 2023 International Conference on Unmanned Aircraft Systems (ICUAS), Warsaw, Poland, 2023, pp. 969-975.
  2. Sanalitro, D. (2022). Aerial Cooperative Manipulation: full pose manipulation in air and in interaction with the environment (Doctoral dissertation, INSA de Toulouse).
  3. D. Sanalitro, H. Savino, M. Tognon, J. Cortés, and  A. Franchi “Full-pose manipulation control of a cable-suspended load with multiple UAVs under uncertainties”. IEEE Robotics and Automation Letters, 2020,5(2), 2185-2191.
  4. C. Gabellieri, M. Tognon, D. Sanalitro and A. Franchi, "Force-Based Pose Regulation of a Cable-Suspended Load Using UAVs with Force Bias,” 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Detroit, MI, USA, 2023, pp. 6920-6926.
  5. N. de Carli, P. Salaris, P. Robuffo Giordano. Multi-Robot Active Sensing for Bearing Formations. MSR 2023 - IEEE International Symposium on Multi-Robot & Multi-Agent Systems, Dec 2023, Boston (MA), United States. pp.1-7.
  6. L. Guanrui, and G. Loianno. "Nonlinear Model Predictive Control for Cooperative Transportation and Manipulation of Cable Suspended Payloads with Multiple Quadrotors." arXiv preprint arXiv:2303.06165 (2023).
  7. A. Ollero, M. Tognon, A. Suarez, D. J. Lee, and A. Franchi. "Past, present, and future of aerial robotic manipulators.’’ IEEE Trans. on Robotics, 2021.
  8. S., Ola, and M. Schwager. "Distributed Model Predictive Control via Separable Optimization in Multi-Agent Networks." IEEE Transactions on Automatic Control (2023).
  9. L. Peric, Brunner, M., Bodie, K., M. Tognon, and Siegwart, R., “Direct Force and Pose NMPC with Multiple Interaction Modes for Aerial Push-and-Slide Operations”, in 2021 IEEE Int. Conf. on Robotics and Automation, Xi’an, China, 2021
Liste des encadrants et encadrantes de thèse

Nom, Prénom
Marco Tognon
Type d'encadrement
Directeur.trice de thèse
Unité de recherche
Centre Inria de l’Université de

Nom, Prénom
Paolo Robuffo Giordano
Type d'encadrement
Unité de recherche
Marco Tognon
Aerial Robotics, Multi-Robot, Manipulation, Control