Technical University of Lisbon (IST)
Loading the player ...
- Offer Profile
- SR-Lisbon is a university
based R&D institution where multidisciplinary advanced research activities
are developed in the areas of Robotics and Information Processing, including
Systems and Control Theory, Signal Processing, Computer Vision,
Optimization, AI and Intelligent Systems, Biomedical Engineering.
Applications include Autonomous Ocean Robotics, Search and Rescue, Mobile
Communications, Multimedia, Satellite Formation, Robotic Aids.
VisLab: Computer and Robot Vision Lab
- The Computer Vision Laboratory - VisLab focuses on the
research and development of tools based on computer vision, mostly for
robotic applications. We are particularly interested in the problems of
active vision, visual based control, motion analysis and segmentation.
DSOR: Dynamical Systems and Ocean Robotics Lab
- The work carried out in the lab is directed towards
furthering knowledge in the general area of dynamical systems theory and
applying it to the design and operation of autonomous marine and aerial
robots. Theoretical areas of research include navigation, guidance, and
control (NGC), mission control, and cooperative control of distributed
autonomous vehicles. Strong cooperation links are being steadily forged with
marine research institutes worldwide as marine science and oceanography
become increasingly dependent on advanced technologies for ocean
exploration. The lab is currently involved in a number of multidisciplinary
projects using advanced robotic systems that include the Infante autonomous
underwater vehicle (AUV), the Delfim autonomous surface craft (ASC), and the
Caravela 2000 autonomous oceanographic vessel.
LaSEEB: Evolutionary Systems and Biomedical Engineering
- The two main areas of this group are: Bio-Inspired
Algorithms (BIA) and Biomedical Engineering (Biomed). Solutions found in
nature inspire the development of search and optimization algorithms as
Swarm Optimization (SO) and Evolutionary Algorithms (EA); modelling and
simulation methodologies as Artificial Life (ALife) and Artificial Immune
Systems (AIS). Neurophysiologic signal and image processing and
Bio-informatics are the main topics of Biomed topic. Human Cognitive states
detection and classification as Wake, Sleep, Drowsiness and Stimulus Related
Responses. Further development is the cross interaction of BIA and Biomed
IRSGroup: Intelligent Robot and Systems Group
- The research and development work carried out at the
Intelligent Robots and Systems Laboratory (IRSLab) is wide in scope. Its
members approach complex systems from a holistic standpoint, rather than
focusing on some of the subsystems. The topic of cooperation (among agents
and/or robots, among robots and humans) arises naturally from this
viewpoint. The historic background of the lab senior researchers has lead us
to use Artificial Intelligence concepts driven by formal approaches that
stem from Systems and Control Theory and from Operations Research. We
further believe it is very important to apply our methodologies to practical
domains, as challenging real-life problems provide richer sources of
SIPG: Signal and Image Processing Group
- Research at the Signal and Image Processing Group (SIPG)
focus on the development of fundamental theory for signal processing on
manifolds, e.g., performance bounds, optimization algorithms, filtering,
manifold learning. Application areas are: wireless communications, including
blind equalization and source separation; underwater, including acoustic
communications and video compression and analysis; time-frequency analysis;
image analysis, including statistical modelling; and video processing,
including motion estimation, tracking of deformable models and inference of
SCTG: Systems and Control Theory Group
- The ISR Systems and Control Theory Group conducts
fundamental research in all aspects of mathematical analyses and design
methodologies, including interdisciplinary research efforts, involving
modern system-theoretic concepts. Current topics under investigation
include: robust multivariable control synthesis, distributed and
decentralized estimation and control systems, hybrid systems, adaptive
control using multiple-model concepts and hierarchical systems. Recent
interdisciplinary investigations relate to the modeling of the human immune
system using hybrid system concepts and modeling of the human vision system
using hierarchical estimation methodologies.
- ISR-CoBot is is an experimental platform for research in
Human-Robot Interaction (HRI). It is a service robot for office environments
designed to perform tasks for users. The research is targeted towards a
robust platform capable of navigating in crowded environments. We aim at
robots that are aware of their own limitations, and are thus capable to
autonomously asking humans for help. This way we expect to get closer to the
goal of a symbiotic interaction between humans and robots.
The platform is based on a customized Nomadic Scout differential drive
platform. Its equipment includes a touchscreen on a laptop for HRI and
computing, an Hokuyo UTM-30LX laser range finder (range 30m), a Kinect RGB-D
camera, and an IP PTZ camera.
(Jan 2014) participation of SocRob@home team, using the ISR-CoBot equipped
with a 5-DoF arm, at RoCKIn Camp 2014, where we obtained the best-in-class
in manipulation award
(Sep 2013) demos of ISR-CoBot at Pavilhão do Conhecimento during European
Rodrigo Ventura -- principal investigator
Miguel Vaz -- MSc student
João Mendes -- technical staf
- Following the success of RAPOSA, the IdMind company
developed a commercial version of RAPOSA, improving it in various ways.
Notably, the rigid chassis of RAPOSA, which eventually ends up being
plastically deformed by frequent shocks, was replaced by semi-flexible
structure, capable of absorbing non-elastical shocks, while significantly
lighter than the original RAPOSA.
ISR acquired a barebones version of this robot, called RAPOSA-NG, and
equipped it with a different set of sensors, following lessons learnt from
previous research with RAPOSA. In particular, it is equipped with:
This equipment was chosen not only to fit better our research interests, but also to aim at the RoboCup Robot Rescue competition. The stereo camera is primarily used jointly with an Head-Mounted Display (HMD) wear by the operator: the stereo images are displayed on the HMD, thus providing depth perception to the operator, while the stereo camera attitude is controlled by the head tracker built-in the HMD. The LRF is being used in one of the following two modes: 2D and 3D mapping. In 2D mapping we assume that the environment is made of vertical walls. However, since we cannot assume horizontal ground, we use a tilt-and-roll motorized mounting to automatically compensate for the robot attitude, such that the LRF scanning plane remains horizontal. An internal IMU measures the attitude of the robot body and controls the mounting servos such that the LRF scanning plane remains horizontal. The IP camera is used for detail inspection: its GUI allows the operator to orient the camera towards a target area and zoom in into a small area of the environment. This is particularly relevant for remote inspection tasks in USAR. The IMU is used both to provide the remote operator with reading of the attitude of the robot, and for automatic localization and mapping of the robot.
stereo camera unit (PointGrey Bumblebee2) on a pan-and-tilt motorized
- a Laser-Range Finder (LRF) sensor on a tilt-and-roll motorized
- an pan-tilt-and-zoom (PTZ) IP camera;
- an Inertial Measurement Unit (IMU).
DSOR-Dynamical Systems and Ocean Robotics Lab
- "The Dynamical Systems and Ocean Robotics group works
towards furthering the knowledge in the general area of dynamical systems
theory and applying it to the design and operation of autonomous marine and
- Title: DEvelopment of Nonlinear Observers
Summary: During the last few decades there has been an extensive
study on the design of observers for nonlinear systems. An observer or
estimator can be defined as a process that provides in real time the
estimate of the state (or some function of it) of the plant from partial and
possibly noisy measurements of the inputs and outputs and inexact knowledge
of the initial condition...
- delfim is a small autonomous catamaran. It has been
designed as a prototype vehicle to proof test the concept of an autonomous
surface craft capable of working in close cooperation with an autonomous
elfim is 3.5 meter long and 2.0 meter wide and is propelled by two electric
motors. Energy is stored in six 12V-55Ah lead acid batteries for propulsion
and four 12V-12Ah lead acid batteries for payload (computers and sensors).
Basic sensores to be installed on board the Catamaran are a DGPS, an
attitude sensor and an echosounder.
- Development of a Long-RangeA
Autonomous Oceanographic Vessel
Worldwide, there has been increasing interest in the analysis of
mesoscale ocean dynamics, that appear to old the key for correct
descriptions and predictions of ocean system behaviour. However, due to the
three dimensional characteristics of the underlying phenomena, the
characterization of mesoscale ocean processes for numerical modeling and
prediction purposes poses formidable challenges to ocean scientists. This
stems from the fact that the space and time scales of the phenomena involved
span ranges from 10km to 300km and a few days to some months, respectively.
With currently available means, it is simply impossible to obtain
oceanographic data with the space and time resolutions required for accurate
ocean modeling. Thus the urgent need to develop advanced technological
systems for cost effective, automatic ocean data acquisition.
- Advanced System Integration for Managing the
Coordinated Operation of Robotic Ocean Vehicles
Three major stumbling blocks have so far prevented demonstrating the
potential applications of Autonomous Underwater Vehicle (AUVs) to demanding
industrial and scientific missions. Namely, i) the lack of reliable
navigation systems, ii) the impossibility of transmitting data at high rates
between the AUV and a support ship at slant range, and iii) the
unavailability of advanced mission control systems that can endow end-users
with the ability to plan, program, and run scientific / industrial missions
at sea, while having access to ocean data in almost real-time so as to
re-direct the AUV mission if required.
- The MORPH project proposes a novel concept of an
underwater robotic system that emerges out of integrating, albeit in a
non-physical manner, different mobile robot-modules with distinct and
complementary resources. It will provide efficient methods to map the
underwater environment with great accuracy in situations that defy existing
technology: namely underwater surveys over rugged terrain and structures
with full 3D complexity, including walls with a negative slope.
Surface module vehicle, MORPH positioning and communications.
Autonomous Surface Vehicle (ASV)
The DELFIMx, first launched in 2007, has been thoroughly tested at sea and
its autonomy and reliability have been instrumental in a number of national
and EU funded projects. It has been used extensively in cooperative missions
with other autonomous surface and underwater vehicles. It can maneuver at
very low speed.
Surface module vehicle, MORPH positioning and communications.
Autonomous Surface Vehicle (ASV)
The MEDUSAS, first launched in 2010, has been thoroughly tested at sea and
its autonomy and reliability have been instrumental in other projects. In
previous missions, equipped with one acoustic modem, it collected data sets
that are currently being used to test the efficacy of single beacon
navigation algorithms. It can maneuver at very low speed. Two of these
vehicles will be available to the project.
- Search & recovery mission (S&R)
Before an underwater mission starts, both the diver and the CADDY system
are "informed" about the mission plan and procedures. The CADDY system will
guide the diver through the mission. Specifically, the scenario is to search
an area in a lawnmowing pattern and recover a specific object. With the
CADDY system, there is no need to perform conventional rope laying on the
sea bottom; instead, the buddy will guide the diver underwater. During the
S&R mission, the autonomous buddy has to
i) follow the predetermined path,
ii) ensure that the diver is following the buddy, i.e. execute cooperative
algorithms with switching leaders, and
iii) keep an appropriate distance to the diver at all times in order to
ensure diver safety and enable interpretation of the symbolic hand gestures
and the diver behaviour.
At any time the diver can stop the mission, change the mission parameters,
or command the buddy to perform compliant tasks. During the validation
scenario, the diver will test the cognitive abilities of the CADDY system,
i.e. behaviour interpretation, symbolic gesture interpretation, and
reactivity of the diver.
Underwater archaeology mission
The diver is led by the CADDY system directly to the place where the
previous diver has stopped with the documentation of the underwater site.
There will be no need for the conventional positioning of frames on the
seabed. When at the exact location, the diver starts with the archaeological
mission (i.e. documentation of the site). While on the site, the diver will
use hand gestures to command the buddy to perform some required tasks such
as take a photo of a part of the sea bottom, make a mosaic of an area,
direct light to a specific part at the sea bottom, etc. to alleviate the
burden on the diver during the execution of a strenuous operation. This set
of tasks will be defined in the project.
At all times during the excavation mission, the autonomous buddy has to
i) attain optimal positioning with respect to the diver in order to ensure
diver safety, enable interpretation of the symbolic hand gestures and
interpretation of the diver behaviour,
ii) accurately interpret commands issued by the diver and comply with the
tasks requested, and
iii) adapt the mission plan according to the diver’s instructions.
The execution of the validation tasks will be assessed by the members of the
CADDY Advisory Board and DAN Europe by referring to the following key
1.speed and success of diver behaviour interpretation;
2.speed and success of the buddy reaction to a change in the mission plan;
3.diver ergonomics – was the distance of the buddy appropriate during the
mission?, was the diver safety area preserved during the mission?;
4.precision of path following (for both the diver and the buddy) during S&R
mission and precision of path following during the descent to the underwater
archaeology site; and
5.precision and compliance of the buddy operations during assistance to the
ATLAS Project: Advances in Terrain-based Localization of
- The Project
The main objective of the current project consists in the development of
geophysical navigation (GN) methods with application to the navigation of
autonomous underwater vehicles (AUVs). This approach to navigation relies
essentially on matching the geophysical data acquired by a vehicle in real
time with pre-existent maps of the area to be surveyed. It is an
alternative, economical approach to dead-reckoning methods based on the
integration of Inertial Navigation Sensor information and Doppler Velocity
Logger data. One of the main advantages of GN is the possibility of
eliminating the drifts that are inherent to the dead-reckoning methods.
Geophysical Navigation is particularly well suited for the navigation of
autonomous vehicles in areas that need to be surveyed repeatedly since the
cost of acquiring the prior maps is diluted along time. As such, GN reveals
high potential of application to oceanography in general and to marine
geophysics in particular. This is an FCT project with the reference PTDC/EEA-ELC/111095/2009.
Introduction and Objectives
A great number of oceanographic missions consist of repeated surveys of the
same area either for successive refinement of feature maps or for monitoring
the evolution of the target environment along time. Examples of such
missions include: marine surveillance in special areas of conservation;
monitoring hydrothermal vents, mud volcanoes, and other geologically active
areas of the sea-floor; habitat mapping; plankton sampling; mine detection
and underwater archeology using sonar imaging and magnetometry; pipeline
inspection; deployment of current profilers for sampling of small-scale
ocean turbulence; inspection of scientific or industrial infrastructures.
Most of the repetitive tasks involved in these surveys can in principle be
efficiently accomplished by Autonomous Underwater Vehicles (AUVs) provided
that a reliable navigation system is available to estimate their position,
velocity, and orientation, and to geo-reference the data acquired. Compared
with other traditional methods, performing data acquisition with AUVs
presents considerable benefits in terms of versatility, safety, operational
costs, and quality of the data acquired. AUVs are ideal vehicles to execute
missions that require 3-dimensional surveys in marine environments.
Furthermore, because AUVs can penetrate the water column and maneuver close
to the seabed in a controlled manner, they can acquire acoustic, magnetic,
and vision data with a resolution that far exceeds that available with
One of the main advantages presented by autonomous underwater vehicles is
the possibility of operating at depths where they are not affected by the
environmental disturbances sensed by surface vessels or even by towed
platforms. The inherent stability of AUVs makes them ideal for the
acquisition of orientation-dependent data. Marine geophysical applications
that can benefit from AUV deployed sensors include vector magnetometry and
magnetic gradiometry. These techniques are routinely used in scientific
research and oil exploration. In the last decades, following pioneer work at
Woods-Hole Oceanographic Institution, near-bottom surveys in deep waters,
including magnetic surveys that traditionally were performed with underwater
towed vehicles, started to be executed by AUVs. In the near future,
motivated by the increasing investment of the oil industry in deep water
surveying for oil exploration and the flurry of scientific activity in the
study of deep ocean habitats and the deployment of underwater laboratories,
we expect to witness a surge of interest in AUVs instrumented for
Conventional AUV navigation methods rely on inertial navigation systems
(INS) or acoustic baselines. High-performance INS are overly expensive units
which are unaffordable to the majority of scientific and commercial
applications. Medium and long baseline systems required by the
aforementioned applications are cumbersome to deploy and pose additional
problems in terms of calibration and maintenance that increase significantly
the complexity of the navigation systems and the costs of operation.
Geophysical Navigation (GN) is an alternative, economical approach that
relies on matching the geophysical data acquired by a vehicle with
pre-existent maps of the area to be surveyed. This method is particularly
well suited for the navigation of autonomous vehicles in areas that need to
be surveyed repeatedly since the cost of acquiring the prior maps is diluted
along time. Previous work by the proponents of this project and other
researchers has shown the potential of geophysical navigation based on sonar
and magnetic data to implement autonomous underwater navigation.
Reportedly, some advanced naval systems employ measurements of geopotential
fields (including gravity and geomagnetism) as navigation aids. It is known
that geophysical navigation technologies have been applied on military
submarines as a mean of compensating the drifts inherent to INS systems.
Sonar bathymetric fix capabilities have been integrated in US Navy
submarines for decades. Given the potential of their application in advanced
weapon systems, in-depth studies involving these technologies remain
withheld from general circulation for reasons of national security. To the
best of our knowledge there is no current implementation of geophysical
navigation systems for routine operations in civilian applications.
Motivated by the above open issues and borrowing from the previous
experience of the research team, we propose in the current project to study,
develop, and validate a number of geophysical navigation techniques
experimentally. The project combines fundamental research with practical
applications and is expected to foster the application of GN methods to
small AUVs, thus contributing to making them affordable and bringing their
potential to bear on the execution of challenging scientific and commercial