Project Leader

Dr. Paolo Stegagno
Phone: +49 7071-601-218
Fax: +49 7071 601-616
Opens window for sending emailpaolo.stegagno[at]tuebingen.mpg.de
Opens external link in new windowWebsite
 
 

Recent Journal Publications

Grabe V, Bülthoff HH, Scaramuzza D und Robuffo Giordano P (Juli-2015) Nonlinear ego-motion estimation from optical flow for online control of a quadrotor UAV International Journal of Robotics Research 34(8) 1114-1135.
Ryll M, Bülthoff HH und Robuffo Giordano P (Februar-2015) A Novel Overactuated Quadrotor Unmanned Aerial Vehicle: Modeling, Control, and Experimental Validation IEEE Transactions on Control Systems Technology 23(2) 540-556.
Zelazo D, Franchi A, Bülthoff HH und Robuffo Giordano P (Januar-2015) Decentralized rigidity maintenance control with range measurements for multi-robot systems International Journal of Robotics Research 34(1) 105-128.
Franchi A, Oriolo G und Stegagno P (September-2013) Mutual Localization in Multi-Robot Systems using Anonymous Relative Measurements International Journal of Robotics Research 32(11) 1302-1322.
pdf
Lee D, Franchi A, Son HI, Ha CS, Bülthoff HH und Robuffo Giordano P (August-2013) Semiautonomous Haptic Teleoperation Control Architecture of Multiple Unmanned Aerial Vehicles IEEE/ASME Transactions on Mechatronics 18(4) 1334-1345.
pdf
Son HI, Franchi A, Chuang LL, Kim J, Bülthoff HH und Robuffo Giordano P (April-2013) Human-Centered Design and Evaluation of Haptic Cueing for Teleoperation of Multiple Mobile Robots IEEE Transactions on Cybernetics 43(2) 597-609.
pdf
Censi A, Franchi A, Marchionni L und Oriolo G (April-2013) Simultaneous Calibration of Odometry and Sensor Parameters for Mobile Robots IEEE Transaction on Robotics 29(2) 475-492.
pdf
Robuffo Giordano P, Franchi A, Secchi C und Bülthoff HH (März-2013) A Passivity-Based Decentralized Strategy for Generalized Connectivity Maintenance International Journal of Robotics Research 32(3) 299-323.
pdf
Franchi A, Secchi C, Son HI, Bülthoff HH und Robuffo Giordano P (Oktober-2012) Bilateral Teleoperation of Groups of Mobile Robots with Time-Varying Topology IEEE Transaction on Robotics 28(5) 1019-1033.
pdf
Franchi A, Masone C, Grabe V, Ryll M, Bülthoff HH und Robuffo Giordano P (Oktober-2012) Modeling and Control of UAV Bearing-Formations with Bilateral High-Level Steering International Journal of Robotics Research 31(12) 1504-1525.
pdf
Franchi A, Secchi C, Ryll M, Bülthoff HH und Robuffo Giordano P (September-2012) Shared Control: Balancing Autonomy and Human Assistance with a Group of Quadrotor UAVs IEEE Robotics & Automation Magazine 19(3) 57-68.
pdf

Export als:
BibTeX, XML, pubman, Edoc, RTF

 

Control of Multiple Robot Systems

Focus of this research project is the control of multiple robot systems. In fact, we believe that the use of many, possibly heterogeneous, cooperating robots is the only way to handle the complexity and the spatial sparseness of the reality.
 
The main research objectives of this project are:
  • Decentralization, i.e., property of
    • being free of any central processing unit;
    • controlling every local action using only locally available information;
    • limiting the amount of computational resources with respect to the number of robots in the group (e.g., FLOPS, memory size, communication bandwidth).
  • Autonomy, i.e., the capability of
    • obtaining information about the environment through onboard sensing and local communication;
    • possessing high mobility;
    • operating for an extended period;
    • having an autonomous decisional ability under dynamic situations;
    • being able to act and interact with a dynamic and sparse environment while quickly adapt to unforeseen changes and being resilient to single-point failures.

Vision-based Formation Control

Click to Enlarge. Two snapshots of an experiment with 3 quadrotor UAVs executing the bearing-only formation controller. On the bottom one can see the onboard-camera images used by the controller to achieve the desired 3D formation.
In this work we address the problem of controlling the motion of a group of UAVs bound to keep a 3D bearing-formation, i.e., defined in terms of only relative angles, as the one that can be measured though onboard cameras. This problem can naturally arise within the context of several multi-robot applications such as, e.g., transportation and coverage. 
 
We define and analyze the concept and properties of bearing-formations, and provide a class of minimally linear sets of bearings sufficient to uniquely define such formations.  
We propose a bearing-only formation controller that resorts to sole bearing measurements, converges almost globally, and maintains bounded inter-agent distances despite the lack of direct metric information.
The controller still leaves the possibility to impose group motions tangent to the current bearing-formation. These can be either autonomously chosen by the robots because of any additional task (e.g., cooperative measurement or exploration), or exploited by a human collaborator. For this latter ‘human-in-the-loop’ case, we propose a multi-master/multi-slave bilateral shared control system providing the co-operator with some suitable force cues informative of the UAV performance. 
 
We validated the proposed theoretical framework by means of experiments with quadrotor UAVs equipped with onboard cameras. We also consider practical limitations, e.g., limited field-of-view.

Decentralized Maintenance of Global Properties

Click to Enlarge. Four snapshots of an experiment with 4 quadrotor UAVs executing the connectivity maintenance control while two humans provide some exogenous inputs aimed. The method is general enough to work with heterogenous robots and both with and without the presence of human collaborators.

We address here the collective motion control problems in which the group of robots is forced to preserve a certain group-property during the task, i.e., to generate an invariant of the controlled group.

As case studies we consider the maintenance of both connectivity and rigidity of the communication network / measurement ensemble, while additional objectives can still be achieved.

Connectivity  is a fundamental property for allowing distributed information exchange among the whole group and therefore to achieve complex collective behaviors. For the connectivity we consider the following cases:

  • continual connectivity, which prescribes that the network must stay connected at any time, and
  • periodical connectivity, which allows the temporary separation in distinct subnetworks enforcing only connection on an average time-basis.

On the other hand the concept of rigidity is related to the observability/measurability of the system and it is also fundamental to keep a certain degree of redundancy in order to ensure robustness to single failures.

The decentralized control strategy is based upon the algebraic graph-theoretic concepts of spectrum of the Laplacian and Rigidity matrix.

We also define a generalized concept of connectivity that is able to integrate many other objectives besides pure connectivity, such as, e.g., collision avoidance  and formation preservation.

Decentralized Multi-target Exploration

Click to Enlarge. Experiments with quadrotor UAVs. Top left: Top view of the experiment. Top right: onboard camera on one of the quadrotors (the one with the red ball in the screenshot in the bottom left and with yellow markings in the bottom right picture). Bottom left: Screenshot of the same exploration in a simulation to visualize connections between quadrotors and some extra information. Bottom right: Side view of the experiment.

In the framework of [RobuffoGiordanoFSB2012] we pursue the problem of exploring a set of important locations in a collective, decentralized,  and efficient way. Possible application fields are, among the others: exploration, surveillance, search and rescue, human-in-the-loop operations, and many more. 

 

The naive combination of an exploration algorithm together with a continual connectivity method leads to either stuck formations or very inefficient exploration. Therefore we present a method where different robots are able to reach the next destination in their list of locations simultaneously as long as this is not fully hindered by the connectivity maintenance objective. The so obtained behavior-based algorithm is guaranteed to be stable and complete. 

 

We perform real experiments with quadrotor UAVs  in order to prove the effectivenes of the proposed collective method and its applicability to shared/human-in-the-loop situation (as required in the real-world scenarios).

 

Visit Thomas Nestmeyer's page for additional information. 

Essential Publications in this Topic

Zelazo D, Franchi A, Allgöwer P, Bülthoff HH und Robuffo Giordano P (Juli-2013) Rigidity Maintenance Control for Multi-Robot Systems In: Robotics: Science and Systems VIII, , 2012 Robotics: Science and Systems Conference, MIT Press, Cambridge, MA, USA, 473-480.
pdfCiteID: ZelazoFABR2012
Robuffo Giordano P, Franchi A, Secchi C und Bülthoff HH (März-2013) A Passivity-Based Decentralized Strategy for Generalized Connectivity Maintenance International Journal of Robotics Research 32(3) 299-323.
pdfCiteID: RobuffoGiordanoFSB2012
Franchi A, Masone C, Grabe V, Ryll M, Bülthoff HH und Robuffo Giordano P (Oktober-2012) Modeling and Control of UAV Bearing-Formations with Bilateral High-Level Steering International Journal of Robotics Research 31(12) 1504-1525.
pdfCiteID: FranchiMGRBR2012
Pasqualetti F, Franchi A und Bullo F (Juni-2012) On Cooperative Patrolling: Optimal Trajectories, Complexity Analysis, and Approximation Algorithms IEEE Transaction on Robotics 28(3) 592-606.
pdfCiteID: PasqualettiFB2011

Export als:
BibTeX, XML, pubman, Edoc, RTF
Last updated: Freitag, 13.02.2015