Assessing Transition of Security Operations in Afghanistan

Assessing Transition of Security Operations in Afghanistan

Assessing Transition of Security Operations in Afghanistan Status Report 18 March 2010 1 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Agenda Problem Statement Methodology System Design Update Values and Metrics Update Preliminary Results Friction Points Earned Value Management 2 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Problem Statement

The goal of the research is to develop a value model that assesses the transition of security lead from the International Security Assistance Force (ISAF) and US Forces Afghanistan (USFOR-A) to the Afghani government and Afghanistan National Security Forces (ANSF) Deputy Director, Force Integration and Training (CJ7) / CSTC-A defined five lines of operation (LOOs) that support the goal of transferring security operations 3 Accelerate ANSF growth Achieve security for the Afghan population Marginalize malign actors Achieve legitimate, responsive, and accountable governance Facilitate community development

Develop metrics and an accompanying decision support tool to measure progress against the five LOOs Stakeholders Force Integration and Training cell of NTM-A/CSTC-A (sponsor) NTM-A/CSTC-A Coalition military leadership U.S. government leadership UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Methodology Project group organized into two subgroups Values and Metrics Research values and metrics Requirements to win a counter-insurgency conflict Assessments of ANSF, security, Afghani government, and community Develop value model with sponsor System Design Development of user interface, input forms, storage, usable

output Integrate values, metrics, and value model from other team into the system 4 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Technical Approach System Design System Input: the quantitative portion of the value model in a standardized survey format, completed by military units System Processing and Storage: completed survey templates are configuration controlled and ingested into data storage. User querying capabilities allow the retrieval of data (by unit and/or AOR and/or date range) to research trends Analysis Output: Condensed and easily understood presentation for decision makers 5 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Concept of Operation Surveys from 5 main military regions CJ7 processes surveys and requests status report

CJ7 Processes Surveys Requests Status Report Military Regions 6 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY System Design Update Past Weeks Progress Input Form Prototype Data Compiler Prototype Query Prototype Output Prototype Way Ahead Obtain weights for Value Model metrics Refine interface and status report requirements

Expand Compiler Capacity 7 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Operational Scenario 8 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Technical Approach Value Model Qualitative Value Model: the identification of an objective hierarchy relating fundamental and means objectives Quantitative Value Model: the articulation of the decision makers preferences towards the attributes, and the means of measuring each attribute V(x) = wivi(xi) where wi = weight of attribute i vi = value of attribute i at score xi 9 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Values and Metrics Update

Near Term Completing value/metric hierarchy without sponsor input Completing weight elicitation without sponsor input Completing input forms Mid Term Provide completed hierarchy with weights and input form to system team 10 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Weights - Theory (1 of 2) Weights represent the relative importance of each parameter vis--vis other parameters at the same node must sum to one at each level under each node Bottom Row weights represent lowest level parameters importance to overall decision; used

for final check with DM Product of the weights on the branch of the tree Also must sum to one Sports Car .7 .3 Performance .1 .35 .25 Braking Acceleration (.7)x(.1) (.7)x(.35) .07 .245 11 Comfort

.3 Handling (.7)x(.25) .175 Top Speed (.7)x(.3) .21 .4 .35 .25 Head Leg Shoulder Room Room Room (.3)x(.4) (.3)x(.35) (.3)x(.25) .075 .12

.105 Bottom Row Weights UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Weights - Theory (2 of 2) First, define Range of Variation (ROV) as the actual range of possible values (from worst to best) for a parameter For example, in our sports car example suppose that, for the cars we are examining, the car with the slowest top speed was 140 mph and car with the highest top speed was 190 mph, with all other cars in between: ROV: 140 190 mph Several methods to elicit weights Direct weights: simply ask DM to provide Swing weights: thought experiment in which DM compares individual attributes directly by imagining (typically) hypothetical outcomes. Robert T. Clemen & Terence Reilly SMARTER: requires only that DM rank order attributes 1 - n Rank Reciprocal: requires only that DM rank order attributes 1 - n Rank Sum: requires only that DM rank order attributes 1 - n 12

UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Swing Weights Example (1 of 2) Assume you have a node with 3 parameters, each with the indicated ROV Annual Income; ROV: $18K - $40K, where higher is better Income Tax Rate; ROV: 1.3% - 8.5%, where lower is better Population; ROV: 250,000 500,000, where higher is better Step 1; Select the one attribute you most want to shift from worst to best Suppose you most prefer to move income from $18K to $40K Step 2; Select the second most desired attribute to change, from worst to best. How important is this to you compared to your first choice ? Suppose you would prefer to move the tax rate from 8.5% to 1.3%, but only half as important as shifting annual income Repeat step 2 for all remaining parameters In our example income is ten times more important than population

UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Swing Weights Example (2 of 2) Summary of DMs feedback on previous slide with K-weight concept: Most important attribute to improve is income: K income = 1 Second most important attribute to improve: Tax Rate Half as important as income therefore Ktax = .5 Kincome Third most important attribute to improve: Population One-tenth as important as income therefore Kpopulation = .1 Kincome Convert to scaled weights that will be inserted the appropriate node of the value model: 14 Kincome Ktax =1

Kpopulation = .1 Kincome = .1 / 1.6 = .0625 weight population = 1 / 1.6 = .625 weight income = .5 Kincome = .5 / 1.6 = .3125 weight tax Column Sums to 1.6 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Preliminary Results Functioning test system using input forms 15 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Preliminary Results Functioning test system using queries 16 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY

Preliminary Results Functioning test system creating an output from queries 17 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Friction Points Distance and interaction of sponsor No face-to-face meetings possible Flow of information is sporadic Use local point of contact for weight elicitation and fabricate unavailable data 18 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Earned Value Management 19 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY

Cost Index 20 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Questions? 21 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY BACKUP SLIDES 22 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Work Breakdown Schedule 23 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY External System Diagrams

24 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY System Analysis Diagram 25 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Functional Architecture 26 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Functional Decomposition 27 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Value Structure (1 of 5)

28 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Value Structure (2 of 5) 29 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Value Structure (3 of 5) 30 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Value Structure (4 of 5) 31 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY

Value Structure (5 of 5) 32 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Metrics and Range of Variation (1 of 7) Accelarate Growth of ANSF: Increase the growth rate of the ANSF (ie obtain desired state more quickly than current plan (when do current plans indicate that ANSF will be at Force Size: What is the size of the ANSF? desired strength/capability?)) Training Capability metric: capacity of facilities metric: #trainers available/#trainers required See ETT/MiTT Availability belwo of #trainers available/required.

metric: number of units fielded ROV: [ANSF = ANA+ ANP]. # troops provided in Afghan Index : ANA: low of 6,000 (in Dec 2003) to high of 95,523 (in Nov 2009); ANP: 95,000 (in Nov 2009). Total ANSF: 6,000 (in 2003, ANP # not available) to 190,523 (ANSF in Nov 2009). metric: average assigned strength Per Afghan Index: as of May 2009 ANP #assigned/#authorized = 99%; No comparable number for ANA. Recruitment Rate metric: monthly recruitment (percent of goal) ROV: [ANSF = ANA+ ANP]. Annual recruitment in Afghan index: ANA: low of 9,671 (Mar 2003 - Feb 2004)

to high of ~34,000 (Mar 2008 - Feb 2009). ANP: 17,474 (Mar 2007 - Feb 2008) and 17,191 (Mar 2008 - Feb 2009). Total ANSF: 9,671 (in 2003/04, ANP # Not Available) to 51, 191 (ANSF in 2008/09). Weapons metric: #rifles on hand / required Number of Units Fielded Average Assigned Strength Availability of Equipment: Do ANSF units have access to proper equipment? Communications Equipment Vehicles Unit Facilities 33

Theoretically Feasible ROV: 0 100% Theoretically Feasible ROV: 0 metric: #MG on hand / required 100% Theoretically Feasible ROV: 0 metric: #radios on hand / required 100% Theoretically Feasible ROV: 0 metric: #vehicles on hand / required 100% Theoretically Feasible ROV: 0 metric: #company HQ avail / required 100% UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Metrics and Range of Variation (2 of 7) Accelarate Growth of ANSF: Increase the growth rate of the ANSF Training and Leadership: What is Unit Rating (ie obtain desired state more quickly the level of training, is there enough Leader to Soldier Ratio leadership for ANSF forces? than current plan (when do current plans indicate that ANSF will be at Advanced Training desired strength/capability?)) metric: # complete/required

Theoretically Feasible ROV: 0 100% Theoretically Feasible ROV: 0 100% ETT/MiTT Availability metric: OH/req ROV: [ANSF = ANA+ ANP]; ANA: 1,062 US Embedded Training Team (ETT) Assigned / 2,391 US ETT Required (44%) (Mar 2008), 1,138 / 2,225 (51%) (Nov 2008), 1,175 / 2,663 (44%) (May 2009) so ROV = 44% - 51%. ANP: 921 US PMT assigned / 2,358 required (39%) (Mar 2008), 886 / 2,375 (37%) (Nov 2008), 1,050 / 2,375 (44%) (May 2009) so ROV = 37% - 44% ANSF Independent Units ROV: per Afghan Index: ANA Units (Total # units projected to be about 120 by 2014): CM1 (0 to 30), CM2 (5 to 30), CM 3 (25 to 40), CM 4(1 to

10) based on Jun 06 - May 09 data. ANP Units (Total # units projected to be about 450 in 2014): CM1 (0 to 30), CM2 (0 to 30), CM3 (5 to 60), metric: # units operating CM4 (300 - 450) based on Feb 08 independently in their own battlespace May 09 data. ANSF Units Conducting Operations at Battalion Level metric:# units conducting operations with/without NATO oversight/support Present for Duty metric: percent present for duty see CM2 above (Battalion level) ROV: Afghan Index, ANA AWOL Rate: 7% (Mar 2007/Feb 2008), 9% (Mar 2008/Feb 2009). ANP AWOL Rate not provided. Present for Duty = 1 - AWOL Rate. metric: reenlistment rate (higher

indicates more confidenc) ROV: Afghan Index, ANA Reenlistment rate: Soldiers 50% (Mar 07/Feb 08) to 57% (Mar 08/Feb09). NCOs: 56% (Mar 07/Feb08) to 63% (Mar 08/Feb09). ANP Reenlistment rate not provided. Staff Training Confidence: Are ANSF forces confident in their ability to egage and defeat insurgents, protect the population? Reenlistments 34 metric: rating in training categories metric: avg leader to soldier ratio for all units metric: # complete/required

UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Metrics and Range of Variation (3 of 7) Confidence: Are ANSF forces confident in their ability to egage and defeat insurgents, protect the Accelarate Growth of ANSF: Increase the growth rate of the ANSF population? (ie obtain desired state more quickly than current plan (when do current plans indicate that ANSF will be at desired strength/capability?)) Kill Ratio metric: #enemy/#ANSF # wins vs # losses Access to Intelligence metric: wins/losses metric: # CO elements that have

access within 10 min metric: # CO elements that have access within 10 min metric: # CO elements with GS support Per Afghan Index: ROV: #ANA killed: 1 (Mar 2007) to 51 (May 2009); #ANP killed: 19 (Feb 2007) to 133 (July 2008). # enemy killed not available (publicaly) Not openly available; SIGACTS ROV: 0-# of ANSF units No data avail ROV: 0-# of ANSF units No data avail ROV: 0-# of ANSF units No data avail Access to Supplies metric: # CO elements with unfilled requests (could specify food, fuel,

and ammunition - Class I,III,V) ROV: 0-# of ANSF units No data avail metric: MiTT assessments ROV: CM4-CM1 or other Data not readily avail on higher HQs Acquisition Operations metric: MiTT assessments ROV: CM4-CM1 or other Data not readily avail on higher HQs Training Plans and Goals metric: MiTT assessments ROV: CM4-CM1 or other Data not readily avail on higher HQs Force Sustainment

metric: MiTT assessments ROV: CM4-CM1 or other Data not readily avail on higher HQs Force Development metric: MiTT assessments ROV: CM4-CM1 or other Data not readily avail on higher HQs Resource Management metric: MiTT assessments ROV: CM4-CM1 or other Data not readily avail on higher HQs Win/Loss Ratios Access to Combat Multipliers Access to MEDEVAC

Access to Fire Support Ministry and HQ Capability: Can Ministry/HQ manage the ANSF effectively? Force Management 35 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Metrics and Range of Variation (4 of 7) Security for Afghan People: Provide safe environment for all factions in the Afghan population Kinetic: Direct, conflict based measures of security. Economic: Indirect measures of security (i.e. higher/lower security impacts these measures)

36 metric: total number of ANSF/total security forces operating in Afghanistan metric: total KIA's from Total KIA's During Preceeding ANSF+ISAF+USFOR-A during past 12 Months 12 months Aghan-on-Afghan Violence metric: number of incidents metric: number of enemy wounded or captured / # killed Wound/Capture vs. Kill Rate % of Security Forces that are from the host nation Voluntary Reporting metric: # of reports in a week Price of Exotic Foods: higher prices reflect increased risk due to difficult (unsecure) transport

metric: average price of selected foodstuff New Business Formations New Construction Capital Flight metric: # of new business permit applications metric: # new construction projects metric: net change in holdings of Aghan banks UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY ROV: 190K (current) to 264K (goal) ROV: (ISAF) 57-571; (ANSF) annual data available Not openly available; SIGACTS Not openly available; SIGACTS Not openly available Need to select foods Not readily available; World Bank has GDP data but only updated

yearly Not readily available $60 million to $2285 million Metrics and Range of Variation (5 of 7) Marginalize Malign Actors: Eliminate influence of malignant individuals/groups # Districts with Functioning Governance metric: number metric: percentage of officials Officials Sleep in District who reside within their district metric: percentage of detainees Detainee Guilt Ratio who turn out to be insurgents metric: percent of captured insurgents with moderate to Captured Insurgent Health serious health problems Media: Television/Newspaper

News/Political Opinion Editorials Internet Access: was forbidden under Taliban Theoretically Feasible ROV: 0 - 365 (However, considering the changing number of districts, we may want this to be a percentage) Theoretically Feasible ROV: 0 - 100% Theoretically Feasible ROV: 0 - 100% Theoretically Feasible ROV: 0 - 100% I believe this is a fallable metric, since it offers equal weight to every media outlet, despite the number of Theoretically Feasible ROV: 0 subscribers. 100% metric: percent sympathetic to insurgents metric: percent of population with internet access Theoretically Feasible ROV: 0 - 100% Hostile to whom? Hostility to U.S. is different from hostility to the Afghan

gov't (neither of which mean support metric: # hostile religious for the Taliban). And, if we can't count services with over 100 attendants attendance, how do we weight them (counting attendence won't effectively? We'd be comparing Religious Leaders happen) apples to oranges. metric: number of districts with shadow governments operating Theoretically Feasible ROV: 0 - 365 Number of Shadow Governments within them 37 UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Metrics and Range of Variation (6 of 7) Accountable Governance: Gov't officials/institutions are honest (not corrupt), fair (do not favor one group over another), legitimate (respected

by the population as the legal authority). Number of New Court Cases Participation in Govt Activities Government Protests Possible ROV: 0 to max number over metric: number of cases opened 2002-2009 for court cases per month Perhaps we can change this to "percent of people who turn to state metric: average number of courts for various cases," as identified Theoretically Feasible ROV: 0 participants in the Afghan Index. 100% metric: number of protests NumberExecution: of OfficialsPerhaps Purchasing Position Budget metric: number

money that is allocated, but not spent, would be a sign of a corrupt metric: money spent / money govt that holds money for itself allocated Formation of Anti-Insurgent LASHKAR Formations metric: number # CIVCAS by ANSF 38 metric: number Possible ROV: 0 to max number over 2002-2009 for government protest of over 100 participants per month Possible ROV: 0 to total number of officials identified Theoretically Feasible ROV: 0 - 1 Possible ROV: 0 to max number over 2002-2009 for formations identified Possible ROV: 0 to max number over

2002-2009 for CIVCAS UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Metrics and Range of Variation (7 of 7) Community Development: Improvement in infrastructure and basic services # Afghan Govt Projects metric: number started metric: dollar value of new projects Electricity Available Community Improvement Medical Care 39 metric: total wattage

metric: hours on per day (from midnight to midnight) metric: polling data metric: number of doctors available per 1,000 inhabitants Possible ROV: Min number to max number over 2002-2009 for projects started that year Possible ROV: Min annual US$ to max annual US$ over 2002-2009 for projects started that year Possible ROV: Min annual MW to Max annual MW over 2002-2009. This can be either total MW or home-generated MW (not including imported) Change this to average number of hours per day We have a good amount of polling data, but none discuss the views Afghan's hold about their own community development. Perhaps we change the wording? Possible ROV: 0 - 1000

UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY MW in 2002 = 243 MW (where 87 is imported), MW in 2007 = 652 (where 167 is imported). Difference in homegeneration: 156 - 485 Theoretically Feasible ROV: 0 - 24 Theoretically Feasible ROV: 0 100% Range of Variation (ROV) aka Range of a value measure Definition: The possible variation of the scores of a value measure -Gregory S. Parnell Important precursor to determining DM value (or utility) function 40 *OR681, GMU UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Weights and Utility Curves (3 of 3)

Elicit utility through lottery or certainty equivalence Weights and utility can be linear, piecewise, exponential, or an S-curve Utility Top Speed (MPH) vs Utility 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 150 160 170

180 190 200 210 220 230 Top Speed (MPH) 41 *OR681, GMU UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY Visit Our Project Website http://mason.gmu.edu/~dugarte/index.html 42

UNCLASSIFIED FOR INSTRUCTIONAL USE ONLY

Recently Viewed Presentations

  • Electrostatic - cpb-us-e1.wpmucdn.com

    Electrostatic - cpb-us-e1.wpmucdn.com

    Three things coulomb's law relate. Charge on object. Distance between them. Force applied between them. ... Force is directly proportional to the size of each charge. When one increase, the other will increase by the same amount. Double 1 charge,...
  • Relationship between conservative dentistry and ...

    Relationship between conservative dentistry and ...

    Relationship between conservative dentistry and periodontology, conservative dentistry and oral surgery. ... technique has been two or more visits 21-years old woman-non successful endodontic treatment tooth N.22,apical clear radiolucency confirming an established lesion bigger than 3mm,it shows ...
  • Criminal Justice: A Brief Introduction, 9/e

    Criminal Justice: A Brief Introduction, 9/e

    PowerPoint Presentation During which policing era did police gain pride in their profession and focus on "traditional" crime-fighting? PowerPoint Presentation Intelligence-led policing is part of which policing era? PowerPoint Presentation This style of policing is marked with a concern for...
  • Chapter Twenty-Six

    Chapter Twenty-Six

    Summary (almost) everything you need to know about micro theory in 30 minutes Production functions Q=f(K,L) Short run: at least one factor fixed Long run: anything can change Average productivity: APL=q/L Marginal productivity: MPL=dq/dL Ave prod. falls when MPL<APL MPL...
  • Building a National Science Digital Library

    Building a National Science Digital Library

    Building Tools and Services on the NDR Dean Krafft, Cornell University [email protected]
  • Summer Safety - Tuscaloosa County School District

    Summer Safety - Tuscaloosa County School District

    Home Alone RULES! Key Rules. Door Rules. Phone Rules. Important Numbers
  • Mental Capacity - Hertfordshire

    Mental Capacity - Hertfordshire

    ability to make a decision and to take action based on that decision. The medical dictionary defines a person with capacity as: "Someone that has sufficient understanding and memory to comprehend the situation a person find themselves and the nature,...
  • 7 Consultation Questions - NHS Sustainable Development Unit

    7 Consultation Questions - NHS Sustainable Development Unit

    7 Consultation Questions. There are seven consultation questions and four of them have two parts. The questions are intentionally high level to ensure the consultation and engagement exercise genuinely provides you with the opportunity to shape the direction of travel...