RESEARCH (In construction)

My general research interests, within the miniature aerial vehicle (MAVs) field, are focused on the theoretical and practical problems arising from the development of autonomous robotic systems. Indeed, the objective of my research is to design fully autonomous unmanned systems that can achieve real-world applications in natural and complex environments by relying only on their onboard intelligent capabilities.

RESEARCH TOPICS

  • Nonlinear Control

  • Multi-Agent Systems



Autonomous Grasping Operational Profile

I am currently developing an autonomous robust aerial grasping system (RAGS) intended to perform fast pick-and-place operations using a hexacopter/quadcopter equipped with a 3DOF/2DOF delta parallel robot.

Challenges:

* Robustness to In-flight manipulation/grasping operations

* Robust and high-rate embedded vision-based object detection strategy/algorithm

* Light and robust onboard manipulator/gripper

* Translational/rotational motion of the rotorcraft

* The aerodynamic profile is more vulnerable to wind disturbances

* Bio-inspired in-motion grasping

Contributions/Propositions

* Airframe adapted for a delta parallel robot (Fig. 4 and 5)

* 40gr carbon-fiber parallel robot (Fig. 5)

* Rotational four-fingered compliant gripper (Fig. 2)

* Vision-based optimal grasping points detection (Fig. 3)

* Task-based Hierarchical Control Strategy


2012 - 2013 / ASSOCIATE POSTDOCTORAL RESEARCHER @ FEMTO-ST Lab (FRANCHE-COMTE University)

My research activities in the field of micro-positioning lies in the framework of the joint project MIM-MAC (High-frequency micro-manipulation). This collaborative project, along with the company Percipio Robotics, focuses on development of microgripper system having an embedded position sensor, where I responsible of proposing and validating experimentally robust control algorithms for the piezoelectric-based 4-DOF microgripper. Likewise, I am conducting the development of a magnetic-based sensor to acquire the micrometric position of the piezoelectric actuators.

My activities within the project are the following:

  • Robust Control of Piezoelectric Cantilevers

  • Development of the magnetic-based position sensor

  • Kalman-based estimation of parasitic nonlinearities (hysteresis and couplings)

1. keyence positioning sensors

2. Piezocantilever (microgripper finger)

3. Hall-effect sensors

4. Magnet


2010 - 2012 / ASSOCIATE POSTDOCTORAL RESEARCHER @ HEUDIASYC Lab - Université de Technologie de Compiègne

As a project researcher at Heudiasyc Lab (UMR-CNRS 7254), at Compiegne, I addressed three subjects:

1. GLMAV (Gun-Launched Micro Air Vehicle) Project.

I was involved in the GLMAV project, headed by the French-German St. Louis Research Institute (ISL) and founded by the Research National Agency (ANR). The GLMAV addresses the fast approaching to a zone of interest to acquire visual information.

  • Navigation and Control algorithms for the aerial robot during the projectile-to-helicopter transition phase

2) Integration of Manipulation Capabilites to Air Rotorcrafts.

I addressed the development of a vehicle concept that combines the 3D-mobility of miniature flying robots with the advantages of robot manipulators, obtaining as a result a hybrid aerial system capable of performing basic prehension (grasping/holding), which is useful to pick-and-place objects from/into ubstructed environments. At this project stage, I am studying the dynamic model of the multi-body vehicle, i.e. quadrotor/manipulator, as well as the nonlinear control strategy to be implemented. The stabilizing algorithm must take into account, either via adaptive or robust techniques, the inherent dynamic couplings between both subsystems

  • Development of a Quadrotor MAV (Video1, Video2)

  • Enhanced Embedded Flight Control System (Rabbit 4300 - 60MHz)

  • Simulink-based Ground Base-station (Video)

  • 30gr Carbon-Fiber Basic Manipulator/Gripper Prototype

  • Euler-Lagrange Modeling of the Multi-body Mechanical System

3. Navigation of Miniature Air Rotorcrafts

Two navigation approaches were proposed for the path-following and trajectory-tracking of quad-rotor. The first approach is based on an alternative bi-rotational model used to follow a vision-based (optical flow) safe flight-path within unstructured environments. Concerning, the trajectory-tracking control, based on the operational capabilities of the quadrotor MAV, an adaptive-robust algorithm is proposed to deal with 2D wind disturbances, either varying or static. Both controllers consider time-scale separation between the rotational (fast dynamics) and translational (slow dynamics) motions

3.1 Vision-Based Reactive Navigation Using a Disturbed Birotational Model &

3.2 Lyapunov-based Robust-Adaptive Control to deal with 2D wind disturbances

2008 - 2010 / Postdoctoral Scholar @ CNRS French-Mexican Joint Research Unit (UMI LAFMIA CNRS 3175)

As a postdoctoral fellow I was directly involved in the technical activation of the French-Mexican Joint Research Unit on Computer Science and Control, where I witnessed the difficulty to launch a new research facility. At this early stage of the postdoctoral fellowship, I had to face new challenges, such as project management and student supervision, which was a valuable experience for my career. Moreover, I was charged to develop two projects:

1) Autonomous Underwater Vehicle

  • Modeling through Euler-Lagrange Formalism

  • Embedded Control Architecture Based on a Rabbit-3400 (30MHz)

  • Homemade IMU

  • Experimental Test: Attitude Stabilization

  • MSc Student supervision

2) Quadtilting Convertible Mini Air Vehicle

  • Modeling through Newton-Euler Formalism

  • Attitude autonomously stabilized

  • Embedded Control Architecture Based on a TI-DSP

  • Experimental Test: Attitude Stabilization

  • MSc supervision

PhD - Université de Technologie de Compiegne (HEUDIASYC Lab) - 2004-2008

The Research carried out during my doctoral studies at the Université de Technologie de Compiegne (HEUDIASYC Lab) the aim was to develop a Tailsitter which is a class of convertible MAVs (cMAVs). The particularity of this robotic system is the combination of forward flight of fixed-wing platforms with VTOL/Hovering capabilities of multirotor helicopters. I conducted theoretical and experimental research regarding the vertical flight regime (rotorcraft mode), since within this operational mode the vehicle is naturally unstable. Moreover, the aerodynamic structure of a tailsitter vehicle is closer to an airplane than a helicopter, this suggests that within vertical-mode such cMAV is not as efficient as in horizontal mode. Thus, the objective was to obtain the best trade-off, in terms performance, between both operational modes. In parallel to the development of the convertible MAV, I proposed and developed the following configurations designs:

(1) Single Tilting Rotor (Video),

(2) Birotor cMAV -Thesis subject (Video)

(3) Triple Tilting Rotor (Video)

  • Experimental Platform

  • Embedded Lyapunov-based Bounded-Input Control

  • Homemade inertial measurement unit (IMU)

  • Embedded Flight Control System based on Rabbit-3400

  • Accelerometer-Gyroscope Fusion via Complementay Filter