lecture_09x - Institute for Signal and Information Processing

forestevanescentΗλεκτρονική - Συσκευές

2 Νοε 2013 (πριν από 3 χρόνια και 8 μήνες)

84 εμφανίσεις

Executive Summery

Amir Harati

What and Why?


Readers (often your boss/manager or potential customers)
can rapidly get acquainted with a large body of the material
without reading it.


It is also called “management summery” because it is used by
managers for decision making.


It is often cited as the most important part of a business plan.


It is a way of marketing of your ideas and plans (Get fund from
investors, Banks, …)


Executive Summery can be written for a accomplished project
or a proposal.


We should always consider potential audiences of any report
or executive summery. ( your advisor or a bank investor ?)

Differences with abstract


Abstracts are shorter.


Abstracts provide an orientation of the
document but you can not replace them with
the document itself.



Abstract is more used in Academia while
executive summery is used in real world.


Building Blocks [1]


Problem Statement: summarize the problem, motivations, why market need this product,
emotional not technical.

P
oor example [2]:

Current gas chromatographs employ NDIR technology, suffering from all the typical problems
associated with narrow frequency spectra analysis.

Good example [2]:

The government needs an extremely accurate remote gas detection system to warn troops
and important facilities of external chemical warfare attacks.


Approaches and Constraints: technical aspects, overview of the design and expressing how
it relates to the desired functionality, requirements/challenges, quantitative
measurements
.

Poor example [2]:

Six nanometer doped metal oxide cerium particles stabilized with oxygenate adsorbents, co
-
dispersed with Nano
-
reverse micelles (water in oil emulsion) provides the basic of our
patent portfolio.

Good example [2]:

The company has developed tested and sold a portable system that detects clouds of toxic
chemicals up to 5 km away from the detector
-

allowing mobile soldiers or sensitive facilities
to detect dangerous chemical clouds from a safe distance


Results: Confirmation that the device meets the requirements, evaluations


Impacts: How this product/plan differs from other competitors?, applications, marketing
(cost, potential consumers,…)






Building Blocks

Poor example for marketing [2]:

Chemical weapons in the hands of terrorists pose a terrible risk to our citizens and soldiers.
Failure to stop a chemical attack would result in untold loss of life.

The chemical detection industry is a $20 billion dollar market and growing at a staggering 5%
annually.

Good example for marketing [2]:

-
Target customers: Large security forces charged with protecting against chemical attack
such as NATO, port/airport security, major city police forces, etc.

-
Target market size: $175 Million

-
Military Force Protection:$50 M

-
Fixed Sites: $50 M

-
Foreign Governments: $75 M


Poor example for competitors [2]:

We have no competition OR None of our competitors use FFT technology instead relying on
the outdated ALP2 algorithms...


Good example for competitors [2] :

Primary Competitors:
Bruker

and General Dynamics

The company's product can detect a wider array of dangerous gases from further away and
with significantly higher accuracy than our competitors.

The company recently defeated both of these competitors in securing a contract with the
government of Singapore, the Asian market leader




Examples (Analyze them)

Car theft is a crime that leaves the owner of a stolen vehicle in a seemingly helpless situation where they feel
that all they can do is wait to see if they ever find out what has happened to their vehicle. Each year,
approximately 1.2 million vehicles are stolen in the United States alone, with a recovery rate of approximately
sixty
-
five percent [3]. The problem is that, once a vehicle is stolen, it can be hidden fairly easily and searching
for a stolen vehicle can be time consuming since it could be moved over great distances. The GPS Tracking
System (GTS) will greatly reduce the time it takes to locate a stolen vehicle by having it automatically report its
position
.


In
order to fulfill its role, the GTS must be able to report its position to the owner of the vehicle. It will be able
to determine if it has been stolen, and if it does come to that conclusion then it will contact the owner’s PC to
report its latitude and longitude and call the owner’s cellular phone. An analog comparator found on the
microcontroller does the monitoring of the siren timers are used to determine how long the signals exist on the
siren wires. Long signals are interpreted as a possible theft, while short signals are interpreted as the owner
using the alarm remote. If the theft detection mechanism is bypassed during the theft of the vehicle, then GTS
will still be useful in the recovery of the vehicle by allowing the owner to poll it for its current location
.


In
order to detect that the vehicle has been stolen, it will require the use of a previously installed car alarm. It
will monitor the activity of the car alarm siren to determine if it is being stolen, and it will be able to
differentiate between the car alarm being armed, disarmed, or set off. The monitoring and controlling will be
performed by an Atmel ATmega128, and will be replaced with the lower cost ATmega64 when it becomes
available. If the alarm is set off and not disabled by the owner’s remote, it will begin tracking the vehicle with a
Garmin GPS receiver. If the vehicle begins to move, then it will use a GSM cellular modem to report its position
to the owner. Disabling the vehicle will be accomplished by opening a relay that will be used to remove power
from the vehicle’s fuel pump
.


Tests
show that we have a workable design. The device detects the car alarm and notifies the owner’s PC. It
sends its location and calls the owner’s cellular phone. The device also currently has the capabilities to disable
a car. The only major challenge left is reducing the cost of the design. The major reasons for the high cost are
the GPS and modem. The rest of the design is very cheap.



The
thing that sets this design apart from other vehicle tracking solutions is that the owner can track their
vehicle themselves without monthly fees, instead of paying a company to track it for them if it should be
stolen. Owning a vehicle with an installed GTS not only will aid in the recovery if it is stolen, but it will also
bring a peace of mind knowing that the recovery chances are also greater.

http://www.ece.msstate.edu/classes/design/ece4542/2003_fall/car_alarm/Deliverables/Design%20Document%20Final2.doc

Examples (Analyze them)

In today’s society, our modern lifestyle and activities require tremendous amounts of energy. This energy is most known as el
ect
ricity, and
the method we use to obtain the energy greatly impacts the environment. The current annual global energy consumption is 4.1x
102
0 J or
13x10
12

W (13 terawatts) and will double by 2050. (International Energy Outlook, 2009) Because the current energy sources are depleti
ng
,
there is a demand for an energy solution. This solution is alternative renewable energies and such energies are wind power, w
hic
h is
harvested through wind turbines; solar energy, which is harvested through solar panels; hydropower, which is harnessed by flo
w o
f water.

Of the renewable energies available, wind energy is rapidly growing because it is an abundant source that is constantly being

re
plenished.
The first reliable information proving the existence of wind energy being harvested through turbines is in 644A.D. in the anc
ien
t area of
Seistan

in the Persian
-
Afghan border (
Hau
, 2005). Since then, turbines have changed dramatically. They are now bigger, more complex, and
produce more energy.

The complexity of turbines is all relative to harnessing as much power available. To be able to extract as much power as pos
sib
le, we have
designed a system which can control the turbine blades angle. It is essentially a system which can control and regulate powe
r o
f the turbine
at different wind velocities and is one of the most important optimization features for a wind turbine system. The efficiency

of

pitch control is
reliant on the response time. It can control rotor power and speed of rotation as well as smoothing out loading variations (
Hau
, 2005).

We analyzed the lift and drag forces that act upon on the poly
-
propylene blades of an existing turbine of height 25cm. The pitc
h control has a
set of bevel gears and a servo motor that turns the turbine blades to optimize the angle of attack. The pitch control can cha
nge

the blades
angles between 0 and 55 degrees in increments of 5 degrees. There is a micro controller which controls the servo motor and w
e h
ave
programmed the micro controller in a way that it can change the blades angle to maximize the power coefficient which is the t
arg
et variable
which we want to control.

A micro controller will be used to control the pitch of the blade angle. The micro controller has a program which collects da
ta
which then
produces a response by pitching the angle of the blade automatically. The pitch control and electrical systems is simplified
to
a linear system
for analysis.

We have designed the pitch control system to optimize the performance of the turbine. Our pitch control has the ability to c
han
ge the blade
angle which can regulate the power by optimizing the power coefficient. The performance of the turbine has been analyzed at d
iff
erent pitch
angles and tested with data collected. From the data we collected, we have concluded that a built in pitch control system is
an
essential
ingredient to optimize the performance of the wind turbine.

As a result of the design, the turbine which currently is manually pitched will automatically control the pitch of the angle
bas
ed on the wind
velocity input it is receiving along with the data that is currently available. Too often, wind turbines shut down occurs du
e t
o damage which
may cause by the high wind speeds. It affects the wind turbines’ ability to harness power that is readily available. The en
erg
y crisis that
society is currently facing is being addressed through renewable energy sources. Wind turbine, besides providing energy from
unl
imited
natural resources, will be more efficient with the implementation of a blade pitch control system.

https://sites.google.com/a/temple.edu/windgen/


Examples (Analyze them)

Liquid Crystal Displays (LCDs) are a very popular way to display information from compact and embedded systems. Systems that

use LCDs are found in classrooms as valuable teaching aids and in company research labs as debugging tools to test developing

products. In both of these situations, the information displayed needs to be easily viewable to all parties involved in test
ing
,
which is not easily achieved with a small LCD screen. There are many large format displays (LFDs) already in production.
However, they cannot be used as direct replacements for LCDs because there are no large format displays in production that
communicate via the Hitachi HD44780 specification, which is the standard for LCD communication. Therefore, a Hitachi
interfaced large format display (HILFD) is needed to provide people performing system tests with an easily readable device to

mimic system LCDs.



The device’s control system will have to spend time processing commands and refreshing the LED arrays and will incorporate a
microcontroller in connection with digital logic to divide the work. The microcontroller will be the element of the system t
hat

is
responsible for conforming to the HD44780 specification, which is the basis for several design constraints. The communicatio
n
standard dictates the response time of the system, the initialization time of the system, the number of wires used for interf
aci
ng,
and the minimum number of characters to be mapped by the system. These standards place requirements on our microcontroller
such as the speed, the amount of RAM available, and the number of I/O pins.



There are other requirements not based on the communication standard. The customer’s needs provide constraint by requiring a

cost efficient price, a relatively small, lightweight package, and power consumption comparable with that of a light bulb.
Technical issues constrain the LED array refresh rate and the design of the LED control circuit.



Since the spring, the HILFD project has overcome the lack of brightness, and the problem has been solved by reducing the duty

cycle driving the LEDs. The HILFD has been packaged but could use an interface module. The HILFD can communicate effectivel
y
using the HD44780 standard, and the groundwork has been laid for a HILFD product line. With manufacturing and technical
design, the HILFD could be propelled into production.



The HILFD will provide an increase in testing facility efficiency. Whether students or technicians, the users will have an e
asi
ly
interfaced, highly visible data display for their system tests. The HILFD will also expand the LFD market into the laborator
y a
nd
testing equipment
industry
by replacing LCDs used at testing benches
.

http://www.ece.msstate.edu/classes/design/ece4542/2003_fall/led_array/docs

Examples (Analyze them)

Approximately 85% of all energy consumption in the United States is produced using non
-
renewable, depleting and
environmentally harmful sources. For wind energy to be a viable alternative, the industry must reduce the cost of operation o
f
windmills. The wind turbine industry is currently plagued with failures of key components, and short life
-
spans of equipment. T
he
component most responsible for downtime of a wind turbine is the gearbox. The gearbox increases the speed of the slow moving
windmill blade to power the generator. Gearbox replacement and lubrication is responsible for approximately 38% of the overal
l
cost of a wind turbine and limits its lifespan to 20 years. The oil circulation system design decreases the financial risk in

fu
nding
wind turbines which paves the way for a surge in growth of the industry.



Our Oil Circulation System will decrease the need for maintenance and extend the life of the gearbox. The oil in the sump wi
ll
be
heated to ensure constant viscosity of oil in the system. An electric pump will send the oil through a series of both magneti
c a
nd
mechanical filters. Pressure censors will monitor the flow of the oil to determine when the filters need to be changed. Oil

wi
ll
then flow through a radiator. Temperature sensors located through the system will determine whether the fan of the radiator
should be turned on to cool the oil. Emergency bypass lines will be places around our devices in case of a breakdown. This wi
ll
ensure that the gears are not damaged due to failure of a component. Nozzles at the end of the system will distribute the oil

to

the places on the gears where it is needed. All the data from the sensors will be fed into the control system. This system wi
ll
control the heater in the sump and the radiator, alert the maintenance persons when the filters need to be changed and direct

th
e
flow of the oil through bypass tubes when there is an emergency.



Our filtration system has exceeded the oil cleanliness ISO 15/13/10 specifications standards set of4406:99. The temperature
sensors contently monitor the changes in oil temperature, switching the heater and radiator fans on and off, to maintain an o
il
temperature of 40
°

C (+/
-

5%). The electric pump is independent of the speed of the windmill so the flow rate is always constant
through the system. The magnetic filter traps ware particles without obstructing the flow of oil. This system can operate for

mo
re
than eight months without any downtime for maintenance while using less than 10kw of output power.



The operational cost of wind energy has decreased by 38%. This has attracted more investors to the market and decreased the
dependence on fossil fuels for energy. This booming industry has increased employment, improved national security and saved
the environment.

https://si tes.googl e.com/a/templ e.edu/oi l
-
ci rcul ati on
-
system
-
for
-
wi ndmi l l
-
gear
-
box/home

Examples (Analyze them)

Wind energy is a clean, independent and sustainable method of producing electricity. Currently, wind is a source that is expe
rie
ncing
incredible growth in the renewable energy sector. This strong growth creates a great investment opportunity as well as a logi
cal

environmental choice. However, wind is a natural phenomenon with tremendous daily, seasonal and annual variability. Turbines
are

used to
convert wind power to electricity, but they require a minimum amount of wind speed to initiate rotation which produces electr
ici
ty. The
variability and minimum requirements of wind power can lead to significant risks of installing turbines that do not generate
eno
ugh electricity
to justify the capital and maintenance expenses. In order for the wind generation market to continue to grow, accurate knowle
dge

of wind
resource potential for a site must be gathered. Wind resources can be measured with anemometers, but placing anemometers over

an

entire
site to account for buildings, open space, man
-
made structures, and other obstacles is unfeasible because of its high cost. It w
ould also be
inaccurate, every location across a site is not encompassed, therefore wind resource potential can only be found by modeling.

Sufficient knowledge of wind resource potential can only come from a model of wind that predicts the wind in all three dimens
ion
s because
wind varies in all three dimensional directions. The proposed model is computer software that utilizes computational fluid dy
nam
ics (CFD) to
make predictions on wind speed, direction, and turbulence. CFD is based on the
Navier
-
Stokes equation which takes into account
acceleration, pressure, and friction.
Navier
-
Stokes outputs the instantaneous state of a fluid in motion, however this is complicated. To
simplify the equation and CFD output, wind modeling is performed in an average state. Data used as input is also averaged, so

Navier
-
Stokes
is modified to include the average and instantaneous states of the fluid. This combined equation is called the Time Averaged
Navier

Stokes
equation (TANS). TANS must be closed using at least one additional equation because TANS contains more variables than equatio
ns.

The
proposed software,
UrbaWind

by
Meteodyn
,
Inc
, uses a single equation to close TANS. Since models must be verified, a representative
anemometer will be set up to determine how accurately the model predicted the wind. The model must also be verified through t
he
extrapolation of wind data measured from other locations. These verifications are constrained to the accuracy of the wind mod
el,

actual wind
measurements, and comparisons to third party sources.

The results of our model established that the Broad Street side of the
McGonigle

building was an adequate site for a wind turbine.
UrbaWind

uses CFD and TANS to successfully model the wind’s speed, direction, turbulence, and pressure on the entire site. The model h
as
been
validated by actual wind measurements to accuracy between 90 and 95 percent. The turbine can use the model to determine place
men
t and
is performing at expectations.
McGonigle

is not the area of fastest wind speed, but other factors such as maintenance, structural stability,
visibility, funding and research accessibility contributed to the selection of
McGonigle’s

roof as the site for the turbine. The model provided
data that demonstrated there would be enough wind for a turbine to generate enough electricity to pay for itself within 8 yea
rs,

given that
the cost of electricity grows at the predicted rate. The wind data was correlated to a power curve for different turbines to
det
ermine the
generation rates. The generations rates were in turn correlated to a financial model another company had created to determine

th
e income
each turbine would produce. The turbine selected is an optimization of available wind power, capital and maintenance costs an
d r
eturn on
investment.

Urban lands have incredible potential for wind power generation. Approximately half the world lives in cities and much power
is
lost in
transmission, locally produced electricity also makes sense. The sheer number of turbines can make an important impact in mee
tin
g global
carbon emissions.
UrbaWind

along with the financial model can provide significant clarification and information to an urban dweller who is
thinking about generating electricity. It is also useful to a campus manager who wants to add a green asset to his electrical

ge
neration
portfolio. If you would like to reduce your carbon footprint,
UrbaWind

software with financial modeling can provide you with the data to
make an informed decision and an incredible investment.


https://sites.google.com/a/temple.edu/urbanwind/

Examples (Analyze them)

Border security, particularly at airports, has become a worldwide concern in the aftermath of the events on 9/11, and it is a

pr
imary thrust of
the Department of Homeland Security and this ITR program. Biometric
-
based systems are expected to be a cornerstone of the U.S. h
omeland
security strategy. The goal of this proposal is to develop an automated
deceit detection

system that determines when a person is being
sufficiently deceptive to warrant further investigation. The long
-
term goal for such a system would be a fully automated system
involving
speech and video inputs and outputs, and an intelligent agent that incorporates sophisticated speech and facial gesture gener
ati
on as well as
understanding. As a first step towards this far
-
reaching goal, we propose automatic processing of speech and video inputs modera
ted by a
human agent who can direct questions based on a number of interpretations provided from their direct surveillance of the subj
ect

and the
automated system. It will be critical that the system not require large amounts of subject
-
specific calibration data or speaker
-
specific
a priori

knowledge. Knowledge of human behavior and subject
-
independent approaches to detection and estimation will be integrated with a
short
amount of calibration data captured from answers to introductory questions.

We propose extending speaker verification technology and developing new model combination frameworks to fuse likelihoods from

ac
oustic
processing, prosody, facial analysis, and external knowledge sources (e.g., specially designed questions) into a single decis
ion
-
making process.
Optimal ways to encode
a

priori

knowledge into conventional statistical modeling approaches will be an important research problem, along
with speaker and response content normalization based on calibration data. We propose a five
-
year plan that consists of three ph
ases. In the
first phase, we will assemble the necessary infrastructure, such as databases and domain knowledge in this new problem area,
and

calibrate
the difficulty of the problem using the first implementation of our proposed system. In the second phase, we will perform bas
ic
algorithm
research on prosody and facial expression analysis, and fusion of these techniques into an integrated decision
-
making system. In

the final
phase of the project, we will evaluate the technology in a limited field test. The proposed research will be one of the first

at
tempts to provide
a rigorous framework for evaluation of such technology for airport security applications.

Intellectual merit:

The national priority area for this proposal is
National and

Homeland Security
. The primary technical focus area is
DMC:
Innovative approaches to the integration of data and analysis systems, including dynamic, data
-
driven models for use in risk
-
ass
essment and
decision
-
making
. The pioneering technical and scientific aspects of our work include (1)

development of resources, including databases and
software, to rigorously study this problem and to benchmark performance; (2)

exploration of prosodic and voice quality cues to d
eceit by
combining traditional events with automatically learned events associated with facial action units;

(3) definition and recogniti
on of new facial
action units for deceit detection; (4)

development of normalization techniques to

enable high performance verification using only small
amounts of calibration data; and (5)

use of risk minimization combined with a dynamic Bayesian network framework to handle multi
ple
asynchronous streams of cues (possibly ambiguous) from audio and video to achieve high performance classification.

Broader impact:

The proposed research offers the potential for the development of biometrics technology of great interest to the
Department of Homeland Security. Major benefits of this proposed research is the ability to increase the effectiveness and to

re
duce the
training time required of human agents, while reducing delays for most travelers. The project will culminate with the develop
men
t of an
application
testbed

that can be used for future research evaluations. If successful, a path to development of this technology for a variety of
similar applications will follow. We also propose conducting four workshops for graduate students and researchers on technolo
gy
related to
this project, and to encourage participation of underrepresented groups at these workshops. All educational materials related

to

these
workshops will be made available on the web to support graduate course work
.


http://www.isip.piconepress.com/publications/courses/msstate/ece_4512/templates/doc/examples/exec_e0.doc

References

[1]http://www.isip.piconepress.com/publications/courses/msstate/ece_4512/templates/doc/ind
ex.html#executive

[2] http://paulgsilva.wordpress.com/2011/05/19/lecture
-
video
-
executive
-
summary
-
good
-
bad
-
ugly/