The Center for Communicating Networked Control Systems

1
The Center for Communicating Networked Control
Systems
Intellectual Vision
Control
Computing
Communication
Information
2
The Center for Communicating Networked Control
Systems

• Center research thrusts
• Coding and modulation for feedback control
• QoS measures for real-time systems
• Access control and transmission control
  protocols for
• maximizing QoS objectives in networked control
  systems
• The controlled formation of ad hoc device
  networks
• Distributed control of sensor/actuator networks
• Uncertainty management
• Distributed and asynchronous control under
  communication
• constraints

3
The Data-rate Theorem was first stated by Center
researchers in 1999 at an ARO workshop
Theorem 1 (Nair, Evans, Tatikonda, Sahai,
Delchamps, Mitter, JB) Consider the
first-order system G(s)b/(s-a). If the
channel capacity R in the feedback loop
satisfies, R gt a log2 e then we may choose a
sampling interval h and a set U of admissible
control values whose cardinality is Ugteah such
that the resulting feedback implementation of
constant gain feedback is boundable.
4
Channel capacity is important when feedback loops
of hard real-time systems are closed over
wireless communication links.
Center researchers were the first to implement
feedback control designs with a Bluetooth
wireless link between the plant and controller.
5
In dimensions higher than 1, classical control
designs cannot typically operate near the
theoretically minimum data-rate
Research is now aimed at source and channel
coding for feedback signals that will allow
control systems to operate near the minimum
data-rate and also over a range of data-rates.
FIGURE A classical lqr design in 2-d finds
that the minimum data-rate is 1.44 times
greeater than the theoretical limit of the
data-rate theorem. At this limit, the only
invariant motion is chaotic on an attractor
consisting of five vertical lines.
6
Controlled motions of multiple robot agents
require high bandwidth position sensing
Typical sensor arrays emphasize relative distance
and bearing measurements
Computer vision
Differential GPS
Sonar
Laser range and bearing sensor
7
New Research Emphasis Freespace Optical
Communications Systems
Early parts of this research involved using high
quality laser optical communications transceivers
(Canobeam from Canon) and mobile robots with
reflectors to establish and maintain optical
links.
8
New Research Emphasis Freespace Optical
Communications Systems
Current research involves a novel modulated
retoreflector device which returns a modulated
signal when a laser beam is incident on the
device.
Light from a dc-laser source is incident on a
corner cube retroreflector, one facet of which
contains a MEMS light modulator. The light which
is reflected returns to the beam source, and is
either coherent (f the light modulator is in it
neutral configuration) or totally defocused.
9
New Research Emphasis Freespace Optical
Communications Systems
Current work in the lab is aimed at integrating
the deformable mirror retroreflector device with
a Mica 2 Mote. This will provide the possibility
of an optical com-munication link with an
RF-enable sensor network.
10
New Research Emphasis Freespace Optical
Communications Systems
An anticipated phase of the project involves
mounting the Mote/ retroreflector device on a
small mobile robot platform. Research will then
be conducted on controlling the robots to
establish and enhance the quality of the optical
comm link.
Returned beam
Incident beam
11
Mobile Free-Space Laser Communication
12
Mobile Free-Space Laser Communication
Modulated retroreflection driven by a modulated
signal from a Mica 2 Mote
Mica 2 Mote generate a modulated signal Modulated
signal is amplified and drives the MEMS light
modulator
Mote generated signal Amplified signal driving
the MEMS light modulator
13
Technology Transfer in 2004 and Beyond
Center for Information and Systems Engineering
(CISE)
http//www.bu.edu/systems
14
Sensor Network Consortium (SNC)

• A university-industry collaborative to drive
  research, development, and technology transfer
• SNC objective To promote interactions that
• Share research innovation and address industry
  RD needs
• Launch joint projects
• Enhance student education

15
Sensor Network Consortium
Arch Rock Corporation BP International Ltd.
Crossbow Technology, Inc. Ember Corporation
The Hartford Honeywell International IBM
Corporation INETCO Systems Limited
Millennial Net, Inc. The Mitre Corporation
Radianse, Inc. SAP Sensicast Systems
Siemens Sun Microsystems Textron Systems
16

17
The Centers Mission to Educate
Photo of students involved with Center-related
research at Boston University. Also depicted in
the new Small Robot Experimental test-bed.
18
Army Laboratory Collaborations

• ARL, Intelligent Optics Laboratory, Dr. Mikhail
  Vorontsov.
• Collaboration with Boston Univ. Precision Optics
  Lab (part of the Center) has resulted in eight
  published papers
• Sound source localization algorithms(developed
  at the
• University of Maryland) have been presented and
  demonstrated to researchers at the Army Research
  Lab (ARL) in Adelphi, MD.
• The Boston University Team has provided hardware
  and
• software for PDA interface/control of robot
  vehicle formation to the Robotic Vehicles group
  (Dr. Stephen Wilkerson) at the Aberdeen Proving
  Ground.