REU Site on Integrated Machine Learning Systems:
Facilities
University of Oklahoma
All students at OU will perform their work within the Artificial
Intelligence and Robotics Super-Laboratory space located in the
the new
Stephenson Research and Technology
Center. This space is shared across several component
laboratories, and includes a shared set of computing resources.
The laboratories include:
- Symbiotic Computing Laboratory (Fagg). This lab includes
the following hardware: Five dual-processor, two quad-processor, and two eight-processor 2.8GHz Pentium-based
Linux boxes; two (modified) P5 gloves (Essential Reality); one
InterSense IS-300 three-axis orientation tracker; one Polhemus Patriot
and two Polhemus Fastrak six-axis position/orientation tracking
systems; one TRACLabs Biclops-based stereo vision system; one robotic
finger testbed consisting of a six-axis load cell (ATI) mounted at the
base of a four-DOF finger; a 1000-node sensor network (a platform for
interactive art); and four two-armed mobile manipulators.
- Interaction Discovery Exploration and Adaptation Lab (McGovern).
The IDEA lab's computing resources currently include
three Mac dual-processors (3.0GHz G4 processors), 6 TB of local (non-RAID) storage, and one mac Mini (Intel-based processor).
- Robotics, Evolution, Adaptation, and Learning (REAL) Lab (Hougen). The REAL Lab
includes the following robots: a FIDO-class planetary rover, six
TXT-based all-terrain robots, and two blimps. These are equipped with
cameras (including a variable-baseline stereo camera and a 9 gram Eyecam),
IR sensors, sonars, shaft-encoders, and SICK laser range-finders. The
electronics assembly room is equipped with professional AC/DC power
supplies, electronic multimeters, and digital oscilloscopes. Computing
facilities include high-end PCs (Linux, MS
Windows, and MacOS).
- OU SuperComputing Center for Education & Research
(OSCER). OSCER has two primary missions: first, to educate OU's
faculty, staff and especially students (both undergraduate and
graduate), not only in the principles underlying High End Computing
(HEC) but also in the practice of effective HEC software design;
second, to collaborate directly with OU researchers in the
application of this knowledge to their specific investigations.
Currently, OSCER maintains several high end computing platforms,
which between them have total peak performance of more than 10
TFLOPs. OSCER has access to high performance networking via
National Lambda Rail and Internet2.
University of New Mexico
- Computer Science Department.
The UNM Computer Science Department has over 200 general-purpose
workstations, running Linux, Windows2000, Mac OS X, and Solaris.
Hardware of special note includes: a Sun Enterprise 4500 with 14
processors and 16 GB of RAM, an HP Itanium cluster with 16 64 bit 733
MHz Itanium processors and 4 GB of RAM, a 1TB Network Appliance file
server, a 220 GB RAID5 linux file server, a 560GB RAID5 linux file
server, a mosix cluster for computer service, an ftp server, and a 10
tape DLT8000 autoloading tape library system for department data
backup. There is a departmental lab available to students, with 25
workstations, primarily linux, as well as several labs dedicated to
specific research projects. The Computer Science Department also
maintains a wireless network (802.llb) in the Farris Engineering
Center for faculty and students. Systems support is provided by two
full time staff and four part-time students. In addition, equipment
for high performance computing and visualization is available through
the High Performance Computing Educational Resource Center
(HPCERC). They have several SMP/MPP cluster systems including the
Roadrunner and Blackbear clusters. In addition, HPCERC maintains an
AccessGrid node for collaborative videoconferencing.
- Multi-Agent, Robotics, Hybrid and Embedded Systems (MARHES)
Laboratory (Fierro) is an interdisciplinary laboratory dedicated to
research and education in hybrid and embedded systems, and robotics.
Computing facilities include Linux and Win32 workstations and laptops.
Robotic hardware includes an unmanned aerial vehicle (UAV), two indoor
Scorpion robots, an 8ft blimp, ten
all-terrain unmanned ground vehicles (UGVs), each equipped with
wireless communication and processing capabilities, a suite of
on-board sensors and computer vision.
MARHES is also equipped with
a state-of-the-art UAV hardware-in-the-loop (HIL) facility consisting
of two Piccolo autopilots (from Cloud Cap), base station,
communication equipment, and software for UAV simulation and
visualization.
Numerous sensors, cameras, frame grabbers (both analog
and digital), embedded computers (PC-104, PDAs), laptops, hardware and
software for data acquisition and control will also be available for
this project.
fagg [[at]] ou.edu
Last modified: Wed Feb 20 00:06:33 2008