Robotics, Intelligent Sensing and Control Lab (RISC)

hesitantdoubtfulΤεχνίτη Νοημοσύνη και Ρομποτική

29 Οκτ 2013 (πριν από 4 χρόνια και 2 μήνες)

107 εμφανίσεις

Robotics, Intelligent
Sensing and Control Lab
(RISC)

University of Bridgeport

Department of Computer Science and Engineering

Robotics, Intelligent Sensing and Control

RISC Laboratory


Faculty, Staff and Students

Faculty: Prof. Tarek Sobh

Staff:


Lab Manager: Abdelshakour Abuzneid


Tech. Assistant: Matanya Elchanani

Students:


Raul Mihali, Gerald Lim, Ossama Abdelfattah,
Wei Zhang, Radesh Kanniganti, Hai
-
Poh Teoh,
Petar Gacesa.

Objectives and Ongoing Projects

Robotics and Prototyping

n
Prototyping and synthesis of
controllers, simulators, and monitors,
calibration of manipulators and
singularity determination for generic
robots.


Real time controlling/simulating/monitoring of
manipulators.


Kinematics and Dynamics hardware for multi
-
degree of freedom manipulators.

Objectives and Ongoing ProjectsRobotics and Prototyping


Concurrent optimal engineering design of
manipulator prototypes.


Component
-
Based Dynamics simulation for
robotics manipulators.


Active kinematic (and Dynamic) calibration of
generic manipulators


Manipulator design based on task specification


Kinematic Optimization of manipulators.


Singularity Determination for manipulators.

Objectives and Ongoing Projects Robotics and Prototyping (cont.)

n
Service robotics (tire
-
changing robots)

n
Web tele
-
operated control of robotic manipulators
(for Distance Learning too).

n
Algorithms for manipulator workspace generation
and visualization in the presence of obstacles.

Objectives and Ongoing Projects

Sensing

n
Precise Reverse Engineering and inspection

n
Feature
-
based reverse engineering and inspection of machine parts.

n
Computation of manufacturing tolerances from sense data

n
Algorithms for uncertainty computation from sense data

n
Unifying tolerances across sensing, design and manufacturing

n
Tolerance representation and determination for inspection and
manufacturing.

n
Parallel architectures for the realization of uncertainty from sensed
data

n
Reverse engineering applications in dentistry.

n
Parallel architectures for robust motion and structure recovery from
uncertainty in sensed data.

n
Active sensing under uncertainty.

Objectives and Ongoing Projects

Hybrid and Autonomous systems

n
Uncertainty modeling, representing, controlling, and observing
interactive robotic agents in unstructured environments.

n
Modeling and verification of distributed control schemes for mobile
robots.

n
Sensor
-
based distributed control schemes (for mobile robots).

n
Discrete event modeling and control of autonomous agents under
uncertainty.

n
Discrete event and hybrid systems in robotics and automation

n
Framework for timed hybrid systems representation, synthesis, and
analysis

Prototyping Environment
for Robot Manipulators

Prof. Tarek Sobh

University of Bridgeport

Department of Computer Science and Engineering

Robotics, Intelligent Sensing and Control

RISC Laboratory


To design a robot manipulator,
the following tasks are required:

n
Specify the tasks and the performance
requirements.

n
Determine the robot configuration and parameters.

n
Select the necessary hardware components.

n
Order the parts.

n
Develop the required software systems (controller,
simulator, etc...).

n
Assemble and test.

The required sub
-
systems for
robot manipulator prototyping:

n
Design

n
Simulation

n
Control

n
Monitoring

n
Hardware selection

n
CAD/CAM modeling

n
Part Ordering

n
Physical assembly and testing

Robot Prototyping Environment

Closed Loop Control

PID Controller Simulator

Interfacing the Robot

Manipulator Workspace
Generation and Visualization
in the Presence of Obstacles


Prof. Tarek Sobh

University of Bridgeport

Department of Computer Science and Engineering

Robotics, Intelligent Sensing and Control

RISC Laboratory


Industrial Inspection and
Reverse Engineering

Prof. Tarek Sobh

University of Bridgeport

Department of Computer Science and Engineering

Robotics, Intelligent Sensing and Control

RISC Laboratory


What is reverse engineering
?

Reconstruction of an object

from sensed information.

Why reverse engineering?

n
Applications:


Legal technicalities.


Unfriendly competition.


Shapes designed off
-
line.


Post
-
design changes.


Pre
-
CAD designs.


Lost or corrupted information.


Isolated working environment.


Medical
.

n
Interesting problem

n
Findings useful.

Closed Loop Reverse
Engineering

A Framework for Intelligent
Inspection and Reverse
Engineering

Recovering 3
-
D Uncertainties
from Sensory Measurements for
Robotics Applications

Prof. Tarek Sobh

University of Bridgeport

Department of Computer Science and Engineering

Robotics, Intelligent Sensing and Control

RISC Laboratory


Propagation of Uncertainty

Refining Image Motion

n
Mechanical limitations

n
Geometrical imitations

Fitting Parabolic Curves

2
-
D Motion Envelopes

Flow Envelopes

3
-
D Event Uncertainty

Tolerancing and Other
Projects

Prof. Tarek Sobh

University of Bridgeport

Department of Computer Science and Engineering

Robotics, Intelligent Sensing and Control

RISC Laboratory


Problem


A unifying framework for
tolerance specification,
synthesis, and analysis across
the domains of industrial
inspection using sensed data,
CAD design, and
manufacturing.

Solution


We guide our sensing strategies
based on the manufacturing
process plans for the parts that are
to be inspected and define,
compute and analyze the tolerances
of the parts based on the
uncertainty in the sensed data
along the different toolpaths of the
sensed part.

Contribution


We believe that our new approach
is the best way to unify tolerances
across sensing, CAD, and CAM, as
it captures the manufacturing
knowledge of the parts to be
inspected, as opposed to just CAD
geometric representations.

Sensing Under Uncertainty
for Mobile Robots

Prof. Tarek Sobh

University of Bridgeport


Department of Computer Science and Engineering


Robotics, Intelligent Sensing and Control

RISC Laboratory


Abstract Sensor Model

We can view the sensory system using three
different levels of abstraction

n
Dumb Sensor
: returns raw data without any
interpretation.

n
Intelligent Sensor
: interprets the raw data
into an event.

n
Controlling sensor
: can issue commands
based on the received events.

3 Levels
o
f Abstraction

Distributed

Control
Architecture

Trajectory of the robot in a
hallway environment

Trajectory of the robot from the initial
to goal point

Trajectory of the robot in the lab
environment

Discrete Event and Hybrid
Systems

Prof. Tarek Sobh

University of Bridgeport

Department of Computer Science and Engineering

Robotics, Intelligent Sensing and Control

RISC Laboratory


The Problem

Hybrid systems that contain a “mix” of:

n
Continuous Parameters and Functions.

n
Discrete Parameters and Functions.

n
Chaotic Behavior.

n
Symbolic Aspects.

Are hard to define, model, analyze,
control, or observe !!

D
iscrete
E
vent
D
ynamic
S
ystems (DEDS) are
dynamic systems (typically asynchronous) in
which state transitions are triggered by the
occurrence of discrete events in the system.


Modified DEDS might be suitable for
representing hybrid systems.

Eventual Goal

Develop the Ultimate Framework and Tools !!

n
Controlling and observing co
-
operating
moving agents (robots).

n
A CMM Controller for sensing tasks.

n
Multimedia Synchronization.

n
Intelligent Sensing (for manufacturing,
autonomous agents, etc...).

n
Hardwiring the framework in hardware
(with Ganesh).

Applications

n
Networks and Communication Protocols

n
Manufacturing (sensing, inspection, and assembly)

n
Economy

n
Robotics (cooperating agents)

n
Highway traffic control

n
Operating systems

n
Concurrency control

n
Scheduling

n
Assembly planning

n
Real
-
Time systems

n
Observation under uncertainty

n
Distributed Systems

Discrete and Hybrid Systems Tool

Discrete and Hybrid Systems Tool

Other Projects

n
Modeling and recovering uncertainty in 3
-
D
structure and motion

n
Dynamics and kinematics generation and analysis
for multi
-
DOF robots

n
Active observation and control of a moving agent
under uncertainty

n
Automation for genetics application

n
Manipulator workspace generation in the presence
of obstacles

n
Turbulent flow analysis using sensors within a
DES framework

THE END