Multi-Medium Vehicles
The Naviator Project
Unmanned Aerial-Underwater Vehicles (UAUVs) arise as a new kind of unmanned systems capable of performing equally well in multiple mediums and seamlessly transitioning between them. The "Naviator" is a multirotor UAUV in an octa-quadcopter configuration developed by Rutgers University and commercialized by SubUAS LLC, and is the first multirotor to effectively perform in air and underwater, with smooth transitions between mediums. As part of my post-doctoral appointment at Rutgers University, I was in charge of the development of new control strategies in order to improve the performance of the "Naviator" in both mediums, while guaranteeing stable transitions between them. Further works were dedicated to the autonomous navigation of the vehicle, being the localization problem one of the main challenges, since GPS signals are rapidly attenuated underwater.
The Naviator, first multirotor multi-medium vehicle.
Quaternion based Singularity-Free Representation
This work focuses in the modeling and trajectory tracking control of a special class of air-underwater vehicle with full torque actuation and a single thrust force directed along the vehicle's vertical axis. In particular, a singularity-free representation was required in order to orient the vehicle in any direction, which becomes critical underwater in order to direct the thrust force in the direction of motion and effectively overcome the increased drag and buoyancy forces. A quaternion based representation is used for this purpose. A hierarchical controller is proposed, where trajectory tracking is accomplished by a Proportional-Integral-Derivative (PID) controller with compensation of the restoring forces. The outer trajectory tracking control loop provides the thrust force and desired orientation. The latter is fed to the inner attitude control loop, where a nonlinear quaternion feedback is employed. A gain scheduling strategy is used to deal with the drastic change in medium density during transitions. The proposed scheme was studied trough numerical simulations, while real time experiments validate the good performance of the system.
Modeling and Control using Hybrid Systems
Modeling and control of a multi-medium unmanned vehicle capable of seamless operation in air or underwater is studied in this work. The multi-medium system is treated as a hybrid system with continuous dynamics while performing in both air and underwater, and discrete jumps in the medium density during the transitions. The continuous dynamics are modeled by the Newton-Euler formalism, taking into account the effects of the buoyancy and drag phenomena, normally neglected in aerial vehicles. A hybrid controller is designed for trajectory tracking considering the full system, including a transition strategy to assure the switching between mediums. Stability analysis for the full system is provided using hybrid Lyapunov and invariance principles. The performance of the control strategy was validated through simulations. Preliminary experimental results validate the proposed configuration while in air, underwater and through transitions.
Autonomous Navigation for Multi-Medium Vehicles
Untethered communication underwater arises as a major challenge, limiting the action of the pilot. Hence, it is desired to perform fully autonomous missions, where precise and reliable localization is required. GPS is the most accepted solution in air, but GPS signals are not available underwater. Meanwhile, acoustic sensors are often used underwater but they are too heavy to be carried in flight. Dead reckoning solutions (velocity or acceleration integration) drift over time without bounds. Visual based navigation presents an interesting solution for some applications.
Autonomous position hold underwater, system description. Computer vision is used to estimate the position of the vehicle w.r.t. a pointcloud of visual features. An EKF is then used to fuse information from the camera with inertial measurements. An embedded computer running ROS is in charge of the pose estimation while an autopilot controls the UAUV.
Visual features captured by the UAUV underwater. Visual based localization appears as an interesting solution in air and underwater, specially for applications such as bridge or dam inspection using multi-medium vehicles.
Videos
Publications
[J2] D. Mercado, M. Maia and J. Diez. Modeling and Control of Unmanned Aerial/Underwater Vehicles using Hybrid Control. Control Engineering Practice, 2018, Vol. 76, pp. 112-122.
[J3] D. Mercado, M. Maia and J. Diez. Aerial-Underwater Systems, a new paradigm in Unmanned Vehicles. Journal of Intelligent & Robotic Systems, Springer, 2018, 10.1007/s10846-018-0820-x.
[C1] M. Maia, D. Mercado and J. Diez. Design and Implementation of Multirotor Aerial-Underwater Vehicles with Experimental Results. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Vancouver, 2017.
[C2] D. Mercado, M. Maia, F. Diez. Aerial-underwater systems, a new paradigm in unmanned vehicles. Unmanned Aircraft Systems (ICUAS), International Conference on, 1690-1695. 2017.