Documente Academic
Documente Profesional
Documente Cultură
Atif Shahzad
_____________________
BE, MECHANICAL ENGINEERING
UNIVERSITY OF ENGINEERING & TECHNOLOGY, TAXILA, PAKISTAN, 2000
EMAIL: atifshahzad@Gmail.com
LINKEDIN: pk.linkedin.com/in/dratifshahzad
PROJECT
MANAGEMENT
COURSE OBJECTIVES
COURSE OBJECTIVES
Learn what project management is and the qualities of an effective
project manager.
Understand the nine knowledge areas of project management and
how they can be applied to your project.
Discover the phases of a project and what deliverables are expected
when.
Identify a project’s key stakeholders.
Understand the different types of business cases and how to create a
Statement of Work.
Learn to be prepared for the unexpected by utilizing risk
management and change control.
Learn how to organize project activities by creating a Work
Breakdown Structure.
Dr. Atif Shahzad
Estimating Project
Times and Costs
Dr. Atif Shahzad
McGraw-Hill/Irwin
Copyright © 2011 by The
McGraw-Hill Companies, Inc. All
Where We Are Now
5–6
Dr. Atif Shahzad
Every project boils down to
Estimating
¤ The process of forecasting or approximating the time
and cost of completing project deliverables.
¤ The task of balancing expectations of stakeholders and
need for control while the project is implemented.
Types of Estimates
¤ Top-down (macro) estimates: analogy, group consensus,
or mathematical relationships
¤ Bottom-up (micro) estimates: estimates of elements
of the work breakdown structure
Dr. Atif Shahzad
Why Estimating Time and Cost Are
Important
5–9
Planning Horizon
Other
Project
(Nonproject)
Duration
Factors
Quality of
Organization Estimates People
Culture
Top-Down Estimates
¤ Are usually are derived from someone who uses experience
and/or information to determine the project duration and total
cost.
¤ Are made by top managers who have little knowledge of the
processes used to complete the project.
Bottom-Up Approach
¤ Can serve as a check on cost elements in the WBS
by rolling up the work packages and associated cost accounts to
major deliverables at the work package level.
Dr. Atif Shahzad
Top-Down versus Bottom-Up
Estimating
5–13
TABLE 5.1
Estimating Projects:
Preferred Approach
5–14
4% Interior decorating
4% Floors laid and finished
Apportion Method of Allocating Project
Costs Using the Work Breakdown Structure
5–25
Dr. Atif Shahzad
FIGURE 5.1
Simplified Basic Function Point Count Process
for a Prospective Project or Deliverable
5–26
Dr. Atif Shahzad
TABLE 5.2
Example: Function Point Count Method
5–27
Dr. Atif Shahzad
TABLE 5.3
PROJECT
MANAGEMENT
ESTIMATING PROJECT TIME & COSTS
BOTTOM UP APPROACH
Bottom-Up Approaches for Estimating
5–29
Project Times and Costs
Template methods
Parametric procedures
applied to specific tasks
Range estimates for
the WBS work packages
Phase estimating: A hybrid
Dr. Atif Shahzad
Support Cost Estimate Worksheet
5–30
Dr. Atif Shahzad
FIGURE 5.2
SB45 Support Cost Estimate Worksheet
Dr. Atif Shahzad
FIGURE 5.2
Phase Estimating over Product Life
Cycle
5–32
Dr. Atif Shahzad
FIGURE 5.3
Top-Down and Bottom-Up Estimates
5–33
Dr. Atif Shahzad
FIGURE 5.4
PROJECT
MANAGEMENT
ESTIMATING PROJECT TIME & COSTS
LEVEL OF DETAIL
Level of Detail
5–35
Direct Costs
¤ These costs are clearly chargeable to a specific work package. Direct costs can be
influenced by the project manager, project team, and individuals implementing the
work package. These costs represent real cash outflows and must be paid as the
project progresses; therefore, direct costs are usually separated from overhead costs.
Lower-level project rollups frequently include only direct costs.
Labor, materials, equipment, and other
Direct (Project) Overhead Costs
¤ Costs incurred that are directly tied to an identifiable project deliverable or work
package.
¤ Direct overhead rates more closely pinpoint which resources of the organization are
being used in the project. Direct project overhead costs can be tied to project
deliverables or work packages.
Salary, rents, supplies, specialized machinery
General and Administrative Overhead Costs
¤ Organization costs indirectly linked to a specific package that are
Dr. Atif Shahzad
FIGURE 5.5
Three Views of Cost
5–41
Dr. Atif Shahzad
FIGURE 5.6
Refining Estimates
5–42
FIGURE 5.7
Key Terms
5–45
TABLE 5.4
Learning Curves Unit Values
5–47
Dr. Atif Shahzad
Learning Curves Unit Values
5–48
Dr. Atif Shahzad
TABLE A5.1
Learning Curves Cumulative Values
5–49
Dr. Atif Shahzad
TABLE A5.2
PROJECT
MANAGEMENT
ESTIMATING PROJECT TIME & COSTS
TYPES OF COSTS
Review
Simple Rules of
5–51
Cost Estimating
1. Assume resources will only be productive for 80 percent of
their time
2. Resources working on multiple projects take longer to
complete tasks because of time lost switching between them
3. People are generally optimistic and often underestimate
how long tasks will take
4. Make use of other people's experiences and your own
5. Get an expert view
Include management time in any estimate
Dr. Atif Shahzad
6.
Simple Rules of
5–52
Cost Estimating
7. Always build in contingency for problem solving, meetings
and other unexpected events
8. Cost each task in the Work Breakdown Structure to arrive
at a total, rather than trying to cost the project as a whole
9. Agree a tolerance with your customer for additional work
that is not yet defined
10. Communicate any assumptions, exclusions or constraints you
have to your customer
11. Provide regular budget statements to your customer,
Dr. Atif Shahzad
¤
¤ The better the project definition, the better the cost estimate
Challenges to Cost Estimate
As spacecraft and mission designs mature, there are many issues and challenges to the
cost estimate, including:
CONCEPTUAL PROJECT
DESIGN DEVELOPMENT OPERATIONS
DEVELOPMENT DEFINITION
A B C D E
PHASE
P
A Analogies , Judgments
R
A System Level CERs As Time Goes By:
M • Tendency to become optimistic
E
T Gen. Subsystem CERs • Tend to get lower level data
R
METHODS
67
¤ CER usually accounts for hardware development and theoretical first unit cost.
For multiple units, the production cost equals the first unit cost times a learning curve factor.
Parametric Cost Estimating
68
$
DDT&E
Y
$
W
$
W
DDT&E
12345678910111213141516171819120212223242526272
123456789101112131415161718191202122232
123456789101112131415161718191202122
123456789101112131415161718191202122232
Production 12345678910111213141516171819120212223242526272
1234567891011121314151617181912021222324252
$ $ 1234567891011121314151617181912021222324252
12345678910111213141516171819120212223242526272
1234567891011121314151617181912021222324252
12345678910111213141516171819120212223242526272
W W 123456789101112131415161718191202122232
Thermal Etc.
Program Specific Input Production
$ $ 12345678910111213141516171819120212223242526272
123456789101112131415161718191202122232
123456789101112131415161718191202122
123456789101112131415161718191202122232
W W 12345678910111213141516171819120212223242526272
• Weight 12345678910111213141516171819120212223242526272
1234567891011121314151617181912021222324252
1234567891011121314151617181912021222324252
• Quantities 12345678910111213141516171819120212223242526272
12345678910111213141516171819120212223242526272
123456789101112131415161718191202122232
(5,32)
• Four data points are available
4
$30
• CER can be derived mathematically using
(2,24) regression analysis
2
(y), Cost
(1,4)
1
$0
0 1 2 3 4 5 6
(x), Weight
458
73
CER Example - Mathematical
$40
(5,32)
4 • Four data points are available
$30
7 • CER can be derived mathematically using
(2,24)
regression analysis
2
(y), Cost
$20
11
• CER based on least squares measure
• “Goodness of fit” is the sum of the squares of
13
the Y axis error
$10 (4,8) • This example compares the eyeball attempt
5 3 with the mathematical look
(1,4)
1
$0
0 1 2 3 4 5 6
(x), Weight
Mathematical Look
Data Summary “Eyeball Try”
Y = 4X +5
Data Point # X Y Data Point # X Y Y Error Y2 Data Point # X Y Y Error Y2
1 1 4 1 1 4 0 0 1 1 9 5 25
2 2 24 2 2 11 13 169 2 2 13 11 121
3 4 8 3 4 25 17 289 3 4 21 13 169
4 5 32 4 5 32 0 0 4 5 25 7 49
458 384
The Best Possible Answer
• Would you prefer a CER or analogy?
• How much do you trust the result? 74
Comparison of Linear / Log-Log Plots
♦ Left side shows the an example CER and data points. Since this is a second order
equation (not a straight line) the relationship is a curve.
♦ A second order equation plots to log-log graph as a straight line and is convenient
for the user, especially when the data range is wide.
$10,000
$800
Sys C
$600
Sys B
Cost
Cost
$400 ($410) $1,000
Sys B Sys C
$200
Sys A
$0
0 200 400 600 800 1000 Sys A
$100
Weight 10 100 1000 10000
Weight
Make Sense?
76
Note: NASA publishes an inflation table (NASA2003_inflation_index.xls)
Use of Complexity Factors
77
Spacecraft / Vehicle Level
$1,000
Cost, (M)
$100
$10
DWT, LBS
KEY
Program Equation Validity Range No of Data Points
Liquid Rocket Engines = 21.364 WT^.5 291 to 18,340 4
Crewed Spacecraft = 19.750 WT^.5 7,000 to 153,552 9
Dr. Atif Shahzad
$10,000
Crewed
DDT&E Flight Flt % of
Weight Cost Unit Cost DDT&E
100 $198.0 $6.4 3.2%
500 $442.0 $19.8 4.5%
$1,000 1,000 $625.0 $32.2 5.2%
5,000 $1,396.0 $99.4 7.1%
10,000 $1,975.0 $162.0 8.2%
20,000 $2,793.0 $262.0 9.4%
50,000 $4,416.0
100,000 $6,245.0
Cost
Earth Uncrewed
DDT&E Flight Flt % of
Weight Cost Unit Cost DDT&E
$10
100 $34.2 $3.8 11.0%
500 $76.6 $11.7 15.0%
1,000 $108.0 $19.0 18.0%
5,000 $242.0 $58.6 24.0%
10,000 $342.0 $95.3 28.0%
$1 20,000 $484.0 $155.0 32.0%
Based on the concept that resources required to produce each additional unit decline as the total
number of units produced increases.
The major premise of learning curves is that each time the product quantity doubles the resources
(labor hours) required to produce the product will reduce by a determined percentage of the prior
quantity resource requirements.
This percentage is referred to as the curve slope. Simply stated, if the curve slope is 90% and it
takes 100 hours to produce the first unit then it will take 90 hours to produce the second unit.
Calculating learning curve (Wright approach):
Y = kxn
Y = production effort, hours/unit or $/unit
k = effort required to manufacture the first unit
x = number of units
n = learning factor = log(percent learning)/log(2); usually 85% for aerospace
Dr. Atif Shahzad
productions
Learning Curve Visual
82
83
7. Consider/include additional costs (wrap factors, reserves, education & outreach, etc.)
8. Document the cost estimate, including data from steps 1-7
Cost estimate includes all aspects of mission effort.
System
IACO
Multi-System
●Non-recurring costs include all costs associated with the design,
development and qualification of a single system. Non-recurring
costs include the breadboard article, engineering model, qualification
Non-Recurring unit and multi-subsystem wraps.
Recurring
● Multi-subsystem wraps are cost related to integrating two or more
Wraps subsystems.
Build / Test B/T ● Recurring costs are those costs associated with the production of
the actual unit(s) to be flown in space. Recurring costs include flight
hardware (the actual unit to be flown85in space) and multi-subsystem
wraps.
Groundrules & Assumptions Checklist (1/2)
86
Assumptions and groundrules are a major element of a cost analysis. Since the results of
the cost analysis are conditional upon each of the assumptions and groundrules being true,
they must be documented as completely as practical. The following is a checklist of the
types of information that should be addressed.
What year dollars the cost results are expressed in, e.g., fiscal year 94$.
Percentages (or approach) used for computing program level wraps: i.e., fee, reserves,
program support, operations Capability Development (OCD), Phase B/Advanced
Development, Agency taxes, Level II Program Management Office.
Production unit quantities, including assumptions regarding spares.
Quantity of development units, prototype or prototype units.
Life cycle cost considerations: mission lifetimes, hardware replacement assumptions, launch
rates, number of flights per year.
Schedule information: Development and production start and stop dates, Phase B
Dr. Atif Shahzad
Authorization to Proceed (ATP), Phase C/D ATP, first flight, Initial Operating Capability
(IOC), time frame for life cycle cost computations, etc.
Groundrules & Assumptions Checklist (1/2)
87
Early scope definition keeps requirement writers from diverging, reduces requirement inconsistencies,
and keeps the BIG PICTURE in view.
It also shortens the time required for requirement writing and rewriting and reduces debates.
The scope definition prevents you from losing sight of important constraints as well as customer needs.
Dr. Atif Shahzad
Groundrules & Assumptions Checklist (2/2)
88
Assumptions and groundrules are a major element of a cost analysis. Since the results
of the cost analysis are conditional upon each of the assumptions and groundrules
being true, they must be documented as completely as practical. The following is a
checklist of the types of information that should be addressed.
Use of existing facilities, modifications to existing facilities, and new facility
requirements.
Cost sharing or joint funding arrangements with other government agencies, if any.
Management concepts, especially if cost credit is taken for change in management
culture, New Ways of Doing Business (NWODB), in-house vs. contract, etc.
Operations concept (e.g., launch vehicle utilized, location of Mission Control Center
(MCC), use of Tracking and Data Relay Satellite System (TDRSS), Deep Space Network
(DSN), or other communication systems, etc.).
Commonality or design heritage assumptions.
Specific items excluded from the cost estimate.
Dr. Atif Shahzad
See also actual page 73 from NASA CEH for other G&A examples
Example of Applying New Ways of Doing Business
to a Cost Proposal
Project X Software Cost
Reconciliation between Phase B Estimates and Phase C/D Proposal
‘87 $ in Millions
8. Other -11
Proposal
48
89
Cost Estimating 89
Selection of Cost Parametric Model
90
Software models exist too; often based on “lines of code” as the independent variable
Sources of Uncertainty in
Parametric Cost Model
H
i • Estimator historical data familiarity
s
t • Independent variable sizing
o
r • Time between / since data points
i
c • Impure data collection
a
l
• Budget Codes
&
• Inflation handling
C
u • WBS Codes
r Affects Cost at:
r • Program nuances (e.g. distributed systems)
e • System Level
n • Accounting for schedule stretches
t • Program Level
• Wraps
probability distribution
+
¤ The cost could be as low as $X, or as high as $Z,
Total Cost
with most likely as $Y
¤ Cost distributions are usually skewed to the right
WBS Element 2
¤ A distribution has positive skew (right-skewed) if
the higher tail is longer +.
Statistically, adding the most likely costs of n WBS
..
elements that are right skewed, yields a result that
can be far less than 50% probable
¤ Often only 10% to 30% probable
Dr. Atif Shahzad
92
Adding Probability to CERs
93 COMBINED COST
MODELING AND CER
TECHNICAL RISK
Cost c
Cost==aa++bX
bXc
COST MODELING risk
Cost
Estimate
The Result of A Cost Risk Analysis
Is Often Depicted As An “S-Curve”
95
100
Estimate at
Dr. Atif Shahzad
Phase A
(very wide
70
distribution)
The intent of Continuous Cost Risk Management
Is to identify and retire risk
so that 70% cost tracks to the left as the project
50
Confidence matures—Historically, it has
more often tracked the other way. But distributions
Level always narrow as project
proceeds.
25
Dr. Atif Shahzad
S-curve or the beginning of Phase B since those times are not the commitment time.
Only the beginning of Phase C is the commitment time.
Confidence Level
Budgeting
Source: NASA/Exploration Systems Mission Directorate, 2007
90%
80%
70%
Confidenc
60%
Level
50%
e
40%
30%
20%
10%
0%
$19.00 $21.00 $23.00 $25.00 $27.00 $29.00 $31.00
Dr. Atif Shahzad
TY $B
PMR 07 Submit 65% Confidence Level 2013 IOC Budget 2015 IOC Budget
Explanation Text to Previous Chart
99
The cost confidence level (CL) curve above is data from the Cx FY07 Program
Manager’s Recommend (PMR) for the ISS IOC scope. The ‘2013 IOC’ point depicts
that the cost associated with the current program content ($23.4B) is at a 35% CL.
Approximately $3B in additional funding is needed to get to the required 65% CL.
Since the budget between now and 2013 is fixed, the only way to obtain the additional
$3B in needed funding is move the schedule to the right.
Based on analysis of the Cx New Obligation Authority (NOA) projection, the IOC date
would need to be moved to 2015 for an additional $3B funding to be available (shown
above as the 2015 IOC point).
Based on this analysis, NASA’s commitment to external stakeholders for ISS IOC is
March 2015 at a 65% confidence level for an estimated cost of $26.4B (real year
dollars).
Internally, the program is managed to the 2013 IOC date with the realization that it is
challenging but that budget reserves (created by additional time) are available to
Dr. Atif Shahzad
Definition: Cost phasing (or spreading) takes the point-estimate derived from a parametric cost model
and spreads it over the project’s schedule, resulting in the project’s annual phasing requirements.
Most cost phasing tools use a beta curve to determine the amount of money to be spent in each year
based on the fraction of the total time that has elapsed.
There are two parameters that determine the shape of the spending curve.
¤ The cost fraction is the fraction of total cost to be spent when 50% of the time is
completed.
¤ The peakedness fraction determines the maximum annual cost.
Cum Cost Fraction = 10T2(1 - T)2(A + BT) + T4(5 - 4T) for 0 ≤T ≤1
Where:
¤ A and B are parameters (with 0 ≤A + B ≤1)
¤ T is fraction of time
¤ A=1, B= 0 gives 81% expended at 50% time
Dr. Atif Shahzad
TIME TIME
Curve 3 Curve 4
TIME TIME
102
Simple Rules of Thumb for Aerospace
Development Projects
CSD (contract start date) to CDR is 50% of project life cycle to first flight unit
delivery to IACO
103
Correct Phasing of Reserves
NO!
YES!
$
Target
Changes
Estimate and
Growth
8 Years
Cost Schedule
Target Estimate $100 M 5 years
Reserve for Changes & Growth $100 M 3 years
Probable $200 M 8 years
104
SUMMARY
COST ESTIMATING
Summary: Cost Estimating
106
Methods for estimating mission costs include parametric cost models, analogy,
and grassroots (or bottoms-up). Certain methods are appropriate based on
where the project is in its life cycle.
Parametric cost models rely on databases of historical mission and spacecraft
data. Model inputs, such as mass, are used to construct cost estimating relationships
(CERs).
Complexity factors are used as an adjustment to a CER to compensate for a
project’s unique features, not accounted for in the CER historical data.
Learning curve is based on the concept that resources required to produce each
additional unit decline as the total number of units produced increases.
Uncertainty in parametric cost models can be estimated using probability
distributions that are summed via Monte Carlo simulation. The S curve is the
cumulative probability distribution coming out of the statistical summing process.
Cost phasing (or spreading) takes the point-estimate derived from a parametric
Dr. Atif Shahzad
cost model and spreads it over the project’s schedule, resulting in the project’s
annual phasing requirements. Most cost phasing tools use a beta curve.
Keys to cost reduction for small satellites
Scale of Project Development and Hardware
¤ Reduced complexity and number of ¤ Using commercial electronics, whenever
interfaces possible
¤ Reduced physical size (light and ¤ Reduced testing and qualification
small) ¤ Extensive software reuse
¤ Fewer functions (specialized, ¤ Miniaturized command & data
subsystems
dedicated mission)
¤ Using existing components and facilities
}
Touch
$30 DDT&E ($128) Non-
90% Prime/Sub Labor
Touch
$0
1 2 3 4 5 6 7 8 9 10
Common Inputs for Parametric Cost Models
Other key parameters
Mass Related
Earth orbital or planetary mission
Satellite dry mass
Design life
Attitude Control Subsystem dry mass
Number of thrusters
Telemetry, Tracking and Command Subsystem Pointing accuracy
mass
Pointing knowledge
Power Subsystem mass
Stabilization type (e.g., spin, 3-axis)
Propulsion Subsystem dry mass
Downlink band (e.g., S-band, X-band)
Thermal Subsystem mass
Beginning of Life (BOL) power
Structure mass End of Life (EOL) power
Notes: Average on-orbit power
Make sure units are consistent with Fuel type (e.g., hydrazine, cold gas)
those of the cost model.
Solar array area
Can use ranges on input variable to Solar array type (e.g., Si. GaAs)
get a spread on cost estimate
Battery Capacity
(high, medium, low).
Battery type (e.g., NiCd, Super NiCd/NiH2)
Need separate model or technique for elements not covered in Small Satellite Cost Model
¤ Concept Development (Phases A&B)
Use wrap factor, as % of Phase C/D cost (usually 3-5%)
¤ Payload(s)
Analogy from similar payloads on previously flown missions, or
Procured cost plus some level of wrap factor
¤ Launch Vehicle and Upper Stages
Contracted purchase price from NASA as part of ELV Services Contract
Follow Discovery Program guidelines
For upper stage, may need to check vendor source
¤ Operations
Analogy from similar operations of previously flown missions, or
Grass-roots estimate, ie, number of people plus facilities costs etc.
¤ Known assets, such as DSN
Get actual services cost from DSN provider tailored to your mission needs
Follow Discovery Program guidelines
¤ Education and Outreach
GRACE mission a good example
Dr. Atif Shahzad
Use of Texas Space Grant Consortium for ideas and associated costs
Analogy
112
Note: for all planetary mission programs, the launch vehicle cost is included in the cost cap.
Cost Estimating Relationships (CERs)
Definition
Equation or graph relating one historical dependent variable (cost) to an independent
variable (weight, power, thrust)
Use
Utilized to make parametric estimates
Steps
1. Select independent variable (e.g. weight)
2. Gather historical cost data and normalize $ (i.e. adjust for inflation)
3. Gather historical values for independent variable values (e.g. weight) and graph cost vs.
independent variable
4. For the plan / proposed system: determine the independent variable and compute the cost
estimate
5. Determine the plan / proposed system complexity factor and adjust the cost estimates
6. Time phase the cost estimate – discussed earlier in this section
113
Cost Estimating 113
COST CONFIDENCE LEVEL
WHY MANY ENGINEERING PROJECTS FAIL
100
Confidence (%)
0
Dr. Atif Shahzad
Cost ($) X
NEAR Actual Costs
Subsystem Actual Cost
in 1997$
Project Management Project Sys Eng Mission Assurance Science Payload Flight System Mission Ops System Launch System
2 01 02 03 04 05 06 07 08
Project Mgmnt Project Sys Eng MA Mgmnt Science Mgmnt P/L Mgmnt Spacecraft Contract Mission Ops Mgmnt Launch Services
3 01.01 02.01 03.01 04.01 05.01 06.00 07.01 08.01
Business Mgmnt Mission & Nav Design System Safety Science Team P/L Sys Eng Flt Sys Mgmnt MOS Sys Eng
01.02 02.02 03.02 04.02 05.02 06.01 07.02
WBS Levels
Risk Mgmnt Project SW Eng Environments Sci Data Support Instrument 1 Flt Sys - Sys Eng Ground Data Sys
01.03 02.03 03.03 04.03 05.03 06.02 07.03
Project Plng Spt Information Systems Reliability Sci Investigatio Instrument N Power Subsys Operations
01.04 02.04 03.04 & Ops Spt 05.04 06.03 07.04
04.04
Review Support Config Mgmnt EEE Parts Eng Sci Environment Common P/L Systems Command & Data S/s MOS V&V
01.05 02.05 03.05 Characterization 05.05 06.04 07.05
04.05
Facilities Planetary Protection HW Q&A Education & Outreach P/L I&T Telecomm Subsys
01.06 02.06 03.06 04.06 05.06 06.05
GN&C Subsys
06.09
Spacecraft Flt SW
06.10
Testbeds
06.11
6 Spacecraft assembly
test & verification
06.12
QUESTIONS
THANK YOU FOR YOUR INTEREST
The project team can create a working software program without
organization or documentation. This is usually called ‘‘spaghetti code.’’
Considering the cost of
the project as delivered, the ‘‘spaghetti coded’’ project will be less costly.
Considering the life cycle cost of the project, however, this approach will be
more costly. This is because the cost of debugging and modifying the soft-
ware after delivery of the project will be more difficult.
Dr. Atif Shahzad
Expected Value ? [Optimistic ? Pessimistic ? (4 ? Most Likely)] / 6
Standard Deviation ? (Pessimistic ? Optimistic) / 6
Dr. Atif Shahzad