Members of the SRV group attended the “Design of applications based on RPAS (Remotely Piloted Aerial Systems)” summer school organized by the Vision4UAV group (UPM) in Madrid, 20-24 June 2016.
On 25-27 May 2016 members of the SRV group attended a testing campaign on-board a bulk carrier in Gdansk (Poland). During testing, several areas of the vessel were covered to check the suitability of the Micro-Aerial Vehicle developed by SRV for the INCASS project. The different areas entailed different operating conditions: higher or shorter heights, higher or lower illumination, uneven floors, etc. The following videos illustrate the inspection tasks performed.
This video contains a summary description of the INCASS MAV prototype developed by members of the SRV group, highlighting the supervised autonomy control approach and a set of behaviours oriented to visual inspection.
The Spanish Association of Artificial Intelligence (AEPIA) honored Laura López Fuentes with the Best Final Degree Project award. The project was selected among the Final Degree Projects presented at the Summer School on Artificial Intelligence (EVIA) held in Carmona (Seville) on June 2016.
Laura López received a degree in Industrial Electronics and Automation from the University of the Balearic Islands on June 2014. In his FDP, completed at the Systems, Robotics and Vision group and supervised by Dr. Oliver-Codina, Laura studied the image vignetting and she designed a new algorithm to reduce this effect and its application to underwater images. Most relevant results of this project were presented at the International Work Conference on Artificial Neural Networks (IWANN) in Palma de Mallorca on June 2015 (see the manuscript here).
Researchers Alberto Ballesteros, Sinisa Derasevic, David Gessner, Francisca Font, Inés Álvarez, Manuel Barranco and Julián Proenza have been awarded with the Best Work in Progress Paper Award (ex aequo) in the 2016 IEEE World Conference on Factory Communication Systems.
The SRV group hosts at the UIB the 7th Technical Meeting of the EU-FP7 project INCASS (Inspection Capabilities for Enhanced Ship Safety) during 22-23 October 2015. The main goal of this project is the development of innovative technological solutions to improve and enhance safety inspections on large tonnage vessels, in order to reduce the risk of the operations and promote maritime security and environment protection.
The meeting will account for the progress on the different project workpackages. Future actions will as well be discussed, such as, in particular, the upcoming field trials for the different robotic platforms developed within the framework of INCASS.
OFFER for Phd EUROWEB Exchange students :
The Systems, Robotics and Vision group (SRV) consists of mostly members of the faculty staff of the Mathematics and Computer Science Department of the UIB (University of the Balearic Islands) and several PhD students. In the last years the group has developed Image Processing and Pattern Recognition methods for specific applications in Underwater Robotics. Updated information can be accessed at: https://srv.uib.es. Concerning Robot Vision, the area in which the research proposal will be carried out, solid results have been obtained by the group related to real-time image processing and understanding algorithms for 3D pose estimation, visual robot guidance and underwater manipulation.
At present, SRV is involved in different funded projects. MERBOTS aims at progressing in the underwater intervention systems development so they can be carried out in a safer way and at a lower cost. The project is being developed under a consortium set up by the universities of Girona, Castellon and Balearic Islands.
Different Robot Vision techniques will be developed under the responsibility of the SRV, ranging from the accurate multimodal 3D reconstruction of the area under study to supervised or unsupervised target detection and tracking. Some more details about these tasks are given next.
Multimodal target detection: We aim at developing the methods necessary to locate the target of the intervention within the environment. To this end, the different sensor data modalities collected during the survey stage and later stored within the map will be analyzed. Different strategies will be considered in order to find matches between map areas and the target, what will lead to a number of potential target detections which will have to be screened in a later step. In any case, the output of these methods will allow end-users to focus their attention on a reduced number of regions of interest instead of the full environment.
Target modelling: Once the target has been detected, and in order for the intervention operation to reach the intended goal(s), appropriate models of the target must be produced from the available sensor data. Using stereo image pairs and/or laser scans, this task will develop methods for building 3D models with the accuracy required by the particular intervention operation, whichever it is. The result will be available for planning the intervention prior to the operation, to be used to track the target in the image stream and to provide the operator with additional input during the operation.
Target tracking: This task is to develop tracking methods that allow the different vehicles involved in the intervention to capture the best sensor data from the target for improved visualization and interaction through the HRI. This will be supplied by both the vehicle performing the manipulation and by an auxiliary vehicle supplying additional views for the intervention. In both cases, the target tracker is to provide feedback to the vehicle controllers to adopt the required pose and, in this way, counteract the constant motion of vehicles, so that the target appears centred in the images. Methods with different computational requirements (and output accuracy) will be developed in order to take advantage or not of the availability of a high-speed communication link between the vehicle and the surface (where powerful computers will be available), i.e. tethered and untethered platforms will be considered separately.
In order to develop part of these tasks framed in a PhD Exchange program, we offer a job position for EUROWEB applicants that want to start and develop his research career in vision for robotics, in a demanding, high technical, friendly and fellowship professional working setting, in the incomparable environment of Mallorca.
We search for motivated PhD students (novel or advanced), from the field of any engineering (preferable electrical, informatics, telecommunications or industrial), with abilities in programming (C, C++), proactive, with an opened mind and with facility and the wish to integrate and work in a highly competitive group, during 1-2 years to develop research tasks.
For further details about the job offer and how to mention it to apply for a EUROWEB grant, please contact Francisco Bonin Font (francisco.bonin@uib.es, +34 971171391). See the official web page of EUROWEB (http://www.mrtc.mdh.se/euroweb/) to see more information about the funded european program and to complete the application.
Seagoing vessels have to undergo regular visual inspections in order to detect the typical defective situations affecting metallic structures, such as cracks and corrosion. These inspections are currently performed by ship surveyors manually at a great cost. To make ship inspections safer and more cost-efficient, this paper presents a Micro-Aerial Vehicle (MAV) intended for visual inspection and based on supervised autonomy. On the one hand, the vehicle is equipped with a vision system that effectively teleports the surveyor from the base station to the areas of the hull that need inspection. On the other hand, the MAV is the result of a complete redesign of a visual inspection-oriented aerial platform that we proposed some years ago, with the aim of introducing the surveyor in the control loop and, in this way, enlarge the range of inspection operations that can robustly be carried out. Another goal is to make the platform as usable as possible for a non-expert. All this has been accomplished by means of the definition of different autonomous functions, including obstacle detection and collision prevention, and extensive use of behavior-based high-level control.
Turbot AUV has been successfully tested under real underwater conditions. The vehicle was surveying autonomously while mapping Posidonia Oceanica meadows at an altitude of three meters.
The trials were conducted in Port of Valldemossa on the 9th of June.