Videos
Loading the player ...
- Offer Profile
- IntRoLab is a research
laboratory pursuing the goal of studying, developing, integrating and using mechatronics and artificial intelligence methodologies for the design of
autonomous and intelligent systems. Research activities involve software and
hardware design of mobile robots, embedded systems and autonomous agents,
and in-situ evaluation. Research conducted at the lab is pragmatic, oriented
toward overcoming the challenges associated with making robots and
intelligent systems usable in real life situations to improve life quality, and also discover how to give intelligence to machines. Application
areas are service robots, all-terrain robots, interactive robots, assistive
robotics, telehealth robotics, automobile and surgical robots
Product Portfolio
MECHATRONIC
DEA
Differential-Elastic Actuator
- The implementation of machines able to precisely control
interaction with its environment begins with the use of actuators specially
designed for that purpose. To that effect, a new compact implementation
design for high performance actuators that are especially adapted for
integration in robotic mechanisms has been developed. This design makes use
of a mechanical differential as central element. Differential coupling
between an intrinsically high impedance transducer and an intrinsically low
impedance spring element provides the same benefits as serial coupling.
However differential coupling allows new interesting design implementations
possibilities, especially for rotational actuators.
DDRA
Dual Differential Rheological Actuator
- Robotic systems are increasingly moving out of factories,
stepping into a dynamic world full of unknowns, where they must interact in
a safe and versatile manner. Traditional actuation schemes, which rely on
position control and stiff actuators, often fail in this new context. There
have been many attempts to modify them by adding a full suite of force and
position sensors and by using new control algorithms but, in most cases, the
naturally high output inertia and the internal transmission nonlinearities
such as friction and backlash remain quite burdensome.
The proposed actuation scheme addresses many of those limitations. The DDRA
uses a differentials mechanism and two magnetorheological brakes coupled to,
for example, an electromagnetic motor. This configuration enables the DDRA
to act as a high bandwidth, very low inertia, very low friction and without
backlash torque source that can be controlled to track any desired
interaction dynamics. The advantages include safety and robustness due to
extreme backdrivability and a lot of versatility in interactions. In a more
traditional context, the actuator’s low inertia, eliminated backlash and
reduced nonlinearities allow for greater accelerations and a more precise
positioning, thus improving productivity and quality.
AZIMUT
Omnidirectional Robotic Platform with
Legs-Tracks-Wheels
- AZIMUT addresses the challenge of making multiple
mechanisms available for locomotion on the same robotic platform. AZIMUT has
four independent articulations that can be wheels (as shown above), legs or
tracks, or a combination of these. By changing the direction of its
articulations, AZIMUT is also capable of moving sideways without changing
its orientation, making it omnidirectional. All these capabilities provide
the robot with the ability to move in tight areas. AZIMUT is designed to be
highly modular, placing for instance the actuators in the articulations so
that the wheels can be easily replaced by leg-track articulations for
all-terrain operations. Stability and compliance of the platform are
enhanced by adding a vertical suspension and using elastic actuators for the
motorized direction of AZIMUT’s articulations. An elastic element is placed
in the actuation mechanism and a sensor is used to measure its deformation,
allowing to sense and control the torque at the actuator’s end. This should
improve robot motion over uneven terrains, making the robot feel the surface
on which it operates. Mechatronic modules, such as the wheel-motor that is
used in all configurations, use a distributed processing architecture with
multiple microcontrollers communicating through shared data buses. AZIMUT's
design provides a rich framework to create a great variety of robots for
indoor and outdoor uses.
CRI -
Roball, Spherical Robot
Tito - Robot Interacting with Children with Autism
- Considering making a robot that can move in a home
environment, filled with all kinds of obstacles, requires particular
locomotion capabilities. A mobile robotic toy for toddlers would have to
move around other toys and objects, and be able to sustain rough interplay
situations. Encapsulating the robot inside a sphere and using this sphere to
make the robot move around in the environment is one solution. The robot,
being spherical, can navigate smoothly through obstacles, and create simple
and appealing interactions with toddlers. The encapsulating shell of the
robot helps protect its fragile electronics. Roball's second prototype was
specifically developed to be a toy and used to study interactions between a
robot and toddlers using quantitative and qualitative evaluation techniques.
Observations confirm that Roball's physical structure and locomotion
dynamics generate interest and various interplay situations influenced by
environmental settings and the child's personality. Roball is currently
being used to see how child interaction can be perceived directly from
onboard navigation sensors.
DEVICES AND TOOLS
RTAB-Map
Real-Time Appearance-Based Mapping
- Loop closure detection is the process involved when
trying to find a match between the current and a previously visited
locations in SLAM (Simultaneous Localization And Mapping). Over time, the
amount of time required to process new observations increases with the size
of the internal map, which may affect real-time processing. RTAB-Map is a
novel real-time loop closure detection approach for large-scale and
long-term SLAM. Our approach is based on efficient memory management to keep
computation time for each new observation under a fixed time limit, thus
respecting real-time limit for long-term operation. Results demonstrate the
approach's adaptability and scalability using one custom data set and four
standard data sets.
AUDIBLE
Sound Source Localization, Separation and Processing
- Artificial auditory system that gives a robot the ability
to locate and track sounds, as well as the possibility of separating
simultaneous sound sources and recognising simultaneous speech. We
demonstrate that it is possible to implement these capabilities using an
array of microphones, without trying to imitate the human auditory system.
The sound source localisation and tracking algorithm uses a steered
beamformer to locate sources, which are then tracked using a multi-source
particle filter. Separation of simultaneous sound source is achieved using a
variant of the Geometric Source Separation (GSS) algorithm, combined with a
multi-source post-filter that further reduces noise, interference and
reverberation. Speech recognition is performed on separated sources, either
directly or by using Missing Feature Theory (MFT) to estimate the
reliability of the speech features. The results obtained show that it is
possible to track up to four simultaneous sound sources, even in noisy and
reverberant environments. Real-time control of the robot following a sound
source is also demonstrated. The sound source separation approach we propose
is able to achieve a 13.7 dB improvement in signal-to-noise ratio compared
to a single microphone when three speakers are present. In these conditions,
the system demonstrates more than 80% accuracy on digit recognition, higher
than most human listeners could obtain in our evaluation when recognising
only one of these sources. All these new capabilities make it possible for
humans to interact more naturally with a mobile robot in real life settings.
The Open Source implementation is called ManyEars and is available at :
http://manyears.sourceforge.net
MARIE
Mobile Autonomous Robot Integrated Environment
- MARIE is a design tool for mobile and autonomous robot
application, designed to facilitate the integration of multiple
heterogeneous software elements. It is a flexible tool based on a
distributed model, thus allowing the realization of an application using one
machine or various networked machines, architectures and platforms. It is
now replaced by ROS from Willow Garage.
Note : MARIE is no longer maintained.
FlowDesigner
Graphical Programming Environ. for Robots
- FlowDesigner is a free (GPL/LGPL) data flow oriented
development environment. It can be used to build complex applications by
combining small, reusable building blocks. In some ways, it is similar to
both Simulink and LabView, but is hardly a clone of either. FlowDesigner
features a RAD GUI with a visual debugger. Although FlowDesigner can be used
as a rapid prototyping tool, it can still be used for building real-time
applications such as audio effects processing. Since FlowDesigner is not
really an interpreted language, it can be quite fast. It is written in C++
and features a plugin mechanism that allows plugins/toolboxes to be easiliy
added.
RobotFlow is a mobile robotics tookit based on the FlowDesigner project. The
visual programming interface provided in the FlowDesigner project will help
people to better visualize & understand what is really happening in the
robot's control loops, sensors, actuators, by using graphical probes and
debugging in real-time. Note : RobotFlow is no longer maintained.
OpenECoSys
- The Open Embedded
Computing Systems (OpenECoSys) project consists in providing free of charge,
open source hardware & software implementations for embedded computing
devices. Initial projet was started from Université de Sherbrooke's IntRoLab
- Intelligent / Interactive / Integrated / Interdisciplinary Robot Lab. Over
time, IntRoLab developed multiple embedded modules for its own mobile robot
platforms. All modules are connected through a shared CAN (Control Area
Network) bus to form a distributed network of sensors and actuators that are
used on advanced platforms such as the AZIMUT3 robot. Most of the embedded
systems are based on Microchip micro-controllers that are inexpensive,
powerful and versatile. Software tools, such as theNetworkViewer was
developed to allow monitoring of multiple internal variable in the
distributed network to facilitate the development of any application.
INTELLIGENT DECISION-MAKING
Autonomous Robot
Social, Intelligent and Autonomous Mobile Robot
- The goal is to design the most advanced robots with the
highest integrated capabilities of perception, action , reasoning and
interaction, operating in natural settings. Integrate many research projects
on the same robotic platform:
Johnny Jr, A Humanoid Robot
Johnny Jr is an interactive robot based on multiple projects (ADE, AUDIBLE,
AZIMUT, HBBA). The robot has different sensors principally configured for
human presence detection. For instance, we use :
- A laser range detector (Hokuyo UTM-30LX) for people's legs on 180 degrees
- A Microsoft Kinect that locates in 3D up to four people in front of the
robot
- An head-mounted camera that detects faces
- An 8 microphones array that localizes sound sources
Johnny Jr can also interact with human in different manners, including:
- Voice
- Facial expressions
- Head movements
- Arm gestures
- Base mobility
HBBA
Hybrid Behavior-Based Architecture
- Our new Hybrid Behavior-Based Architecture (HBBA)
combines our latest research in autonomous mobile robots control
architectures and is the logical evolution of our Motivated Behavioral
Architecture (MBA). It currently powers the mind of our humanoid robot,
Johnny 0.
Our decisional structure remains heavily distributed. Motivation modules
provides the system with high-level desires. The Intention Workspace acts as
a blackboard - a sharing space for motivation modules to post their desires.
The Intention Translator takes these desires and translate them into a
tangible intention for the controlled robot. This means our motivation
modules are loosely coupled with the actual robot, making them portable
between different instances of our architecture. Selective attention
mechanisms implemented within the translator try to make the most of our
platforms' limited resources by selecting competing perceptual and
behavioral strategies in line with the current situation. The Egosphere
surveys perceptual events and builds a more manageable abstraction of every
sensorial inputs provided by our hardware. The Emotional Subsystem regulates
desires' intensities according to the simulated mood of the robot, which
varies in relation to the current intention accomplishment.
Its current implementation is a set of reusable ROS packages, mostly written
in C++.
DCD
Collaborative Driving Systems
- To eventually have automated vehicles operate in
platoons, it is necessary to study what information each vehicle must have
and to whom it must communicate for safe and efficient maneuvering in all
possible conditions. By emulating platoons using a group of mobile robots,
we demonstrate the feasibility of maneuvers (such as entering, exiting and
recuperating from an accident) using different distributed coordination
strategies. The coordination strategies studied range from no communication
to unidirectional or bidirectional exchanges between vehicles, and to fully
centralized decision by the leading vehicle. Instead of assuming that the
platoon leader or all vehicles globally monitor what is going on, only the
vehicles involved in a particular maneuver are concerned, distributing
decisions locally amongst the platoon. Experimental trials using robots
having limited and directional perception of other, using vision and
obstacle avoidance sensing were conducted. Results confirm the feasibility
of the coordination strategies in different conditions, and various uses of
communicated information to compensate for sensing limitations.
INTERFACES AND INTERACTION
TRInterface
Egocentric and Exocentric Teleoperation
- Following the original Telerobot project, we began
development on a novel 3D interface for teleoperated navigation tasks.
This interface combines, in real time:
- An extrusion of SLAM-built 2D map of the environment
- A laser-based surface projection of a 2D video feed
- A 3D projection of a colored point cloud built from a stereoscopic camera
- A CAD-based 3D model of the robot
With this interface, the user can seamlessly transition from egocentric to
exocentric viewpoints by moving the virtual camera around the controlled
robot. As in modern 3D third-person videogames, the user is able to set its
viewpoint to best suit the task at hand, like a top-down view to navigate
tight spaces or a straight-ahead view to communicate with people.
Future work
A new, ROS-compatible and open source implementation that will take
advantage of modern sensors like the Kinect is currently being worked on.
Teletrauma
Telecoaching in Emergency Rooms
- Purpose: To compensate for the shortage of emergency
specialists and to maintain quality of medical care in remote regions.
Methods: Develop and validate the concept of real-time coaching using a
robotized camera system ceiling-mounted above a stretcher. Exploit a final
prototype to evaluate its impact on surgical procedures under controlled
conditions.
Results : Preliminary results demonstrate: i) the system allows a remote
practitioner to easily view all surgical procedures used in emergency
room, and ii) a specialist can coach a nonspecialist to perform a surgical
procedure using proper kinesiologic gestual with many individuals
surrounding the stretcher. Conclusion: The system could provide help on
demand and improve the level of services in trauma medicine for the
population of regions where trauma surgeons are unavailable. The system
could also decrease medical cost by providing remote support on surgical
procedures used to stabilize unstable polytraumatized patients at their
arrival in the trauma room.
Telerobot
Mobile Robot for Home Telepresence
- Telehealth assistive technologies for homes constitute a
very promising avenue to decrease load on the health care system, to reduce
hospitalization period and to improve quality of life. Teleoperated from
adistant location, a mobile robot can become a beneficial tool in health
applications. However, design issues related to such systems are broad and
mostly unexplored (e.g., locomotion and navigation in-home settings, remote
interaction and patient acceptability, evaluation of clinical needs and
their integration into health care information systems). Designing a safe
and effective robotic system for in-home teleassistance requires taking into
consideration the complexities of having novice users remotely navigate a
mobile robot in a home environment while they interact with patients.
Interdisciplinary and exploratory design methodology is adopted to develop a
telepresence assistive mobile robot for homecare assistance of elderly
people. Preliminary studies using robots, focus groups and interviews
allowed us to derive preliminary specifications to design a new mobile
robotic system named Telerobot. Telerobot’s locomotion mechanism provides
improved mobility when moving on uneven surfaces, helping to provide stable
video feed to the user. Its control system is implemented for safe
teleoperation. A study involving 10 rehabilitation professionals confirms
that the system is usable in home environments. Analysis of teleoperation
strategies used by novice teleoperators suggest that it is essential in a
home environment that the teleoperation interface provides the user with a
visual feedback of the objects surrounding the robot, their distances
relative to the robot and the size of the robot in the environment. Enhanced
user interfaces to augment the operator’s perception of the environment were
elaborated and tested in controlled conditions. These experiments are
conducted with the objective of coming up with a complete, efficient and
usable in-home teleassistance mobile robotic system.
CRI
Ecosystemic Studies of Child-Robot Interaction
- Considering making a robot that can move in a home
environment, filled with all kinds of obstacles, requires particular
locomotion capabilities. A mobile robotic toy for toddlers would have to
move around other toys and objects, and be able to sustain rough interplay
situations. Encapsulating the robot inside a sphere and using this sphere to
make the robot move around in the environment is one solution. The robot,
being spherical, can navigate smoothly through obstacles, and create simple
and appealing interactions with toddlers. The encapsulating shell of the
robot helps protect its fragile electronics. Roball's second prototype was
specifically developed to be a toy and used to study interactions between a
robot and toddlers using quantitative and qualitative evaluation techniques.
Observations confirm that Roball's physical structure and locomotion
dynamics generate interest and various interplay situations influenced by
environmental settings and the child's personality. Roball is currently
being used to see how child interaction can be perceived directly from
onboard navigation sensors.