Base Camp Quality of Life Standardization and

Improvement
Patrick K. Kuiper Stephan E. Kolitz Vahid Tarokh
United States Military Academy Draper Harvard University
West Point, NY 10997 Cambridge, MA 02139 Cambridge, MA 02138
Email: [Link]@[Link] Email: kolitz@[Link] Email: vahid@[Link]

Abstract—The United States (US) Army has over 66,000 maintain a force that can project power with agility, the US
soldiers engaged in contingency operations across the world. Army must establish efficient base camp design doctrine. [3]
Current budgetary constraints and an uncertain global security We provide a methodology to maintain soldier mission
environment require these operations to be executed as efficiently
as possible. Base camps are the secured areas where soldiers live effectiveness under resource constraints by recommending
when deployed to contingency operations. Base camps impose base camp design doctrine that improves quality of life (QOL)
a significant financial and tactical burden during contingency while using less resources. QOL on US Army base camps
operations and sub-optimal soldier quality of life decisions have depends upon the essential services provided to soldiers to
significantly contributed to costs. Quality of life (QOL) refers to maintain their mission effectiveness in contingency operations.
the non-security and non-mission related services that directly
sustain the mission effectiveness of soldiers. Current US Army [5] The principal resources that constrain QOL on base camps
base camp tactics, techniques, and procedures (TTPs) do not suf- are fuel, water, waste water, and solid waste. [20] These four
ficiently specify QOL services, and more detailed doctrine should constraining resources represent a fundamental subset of the
be developed to support combat units executing contingency resources that limit the US Armys ability to project power in
operations. In this investigation we employ quantitative methods distributed global contingency operations.
to select decisions that improve QOL and inform doctrine.
We leverage a QOL function and resource consumption data Currently subject matter experts employ anecdotal experi-
developed by US Army Natick Soldier Research, Development ences to inform Tactics, Techniques, and Procedures (TTPs)
and Engineering Center’s (Natick Labs) to build a model that that are used for base camp design standards. These TTPs are
improves QOL under the constraints of four fundamental re- provided in regional command Standard Operating Procedure
sources: fuel, water, waste water, and solid waste. We employ (SOP) documents; however, the US Army has not established
a mixed integer linear program modeling approach and execute
sensitivity analysis to evaluate the strength of our results. Our base camp design doctrine implementing standard QOL ser-
final model is formulated as a chance constraint optimization to vices across all organizations. [12], [13] The lack of doctrine
address the uncertainty associated with resource availability in for base camp design, and resulting QOL, results in inefficient
contingency operations. Our results provide QOL decisions that resource consumption and high costs measured in monetary
reduce resource consumption while maintaining an equivalent expense and reduced operational capabilities. [3], [18]
QOL level when compared to current TTPs. The model provides
quantitative rigor, informing decision makers of specific base The methodology we provide optimizes QOL on US Army
camp design principles for the development of doctrine. base camps, accounting for interactions among dozens of
QOL services, constraining resources, and uncertainty. Current
practice employing anecdotal subject matter expert decision
I. I NTRODUCTION making alone to dictate TTPs does not rigorously account for
A. Problem Description and Definition the complex interactions and trade-offs present when designing
The United States of America’s (US) strategic policy re- and operating a base camp in contingency operations.
quires that the US Army execute sustained contingency op- B. Modeling Approach and Limitations
erations across the world. The objectives of these contin-
gency operations span the range from high-intensity lethal We start to address this problem by leveraging research
engagements to disaster and humanitarian relief; however, all from US Army Natick Soldier Research, Development and
these require a globally distributed troop presence. [27] This Engineering Center’s (Natick Labs) Sustainable and Logis-
presence is enabled by base camps - areas where soldiers tics Basing Science and Technology Objective Demonstration
live while conducting contingency operations, and where their (SLB-STO-D). Natick Labs SLB-STO-D has built a QOL
mission effectiveness is maintained or improved. [6] Coupled function that measures soldier QOL on US Army base camps
with operational requirements are the contemporary budgetary based upon QOL decisions made when designing the camps.
constraints placed upon the US Army - while demand for [2], [5] These QOL decisions include hygiene services, food
the Army grows, available resources are being reduced. To quality, heating and cooling of living space, and Morale
Welfare Recreation (MWR) options - all fundamental services
that maintain solider mission effectiveness in contingency
U.S. Government work not protected by U.S. copyright
operations. Employing additional research that was executed [26], [33] Our formulation of the resource constraints on base
at Natick Labs SLB-STO-D we measure each QOL decision’s camps was informed by investigations from Anderson and
contribution to the consumption of fuel, water, and subsequent motivated by the work of Marsh and the Noblis Corporation.
production of waste water, and solid waste. Using all this [1], [3], [16], [18], [20]
information we formulate a deterministic Base Camps Mixed The parameters of the BCMIP were developed based on
Integer Program (BCMIP). We reformulate the BCMIP to technical reports completed by Natick Labs SLB-STO-D. Our
include chance constraints (BCMIP-CC) in order to account QOL objective function was developed from referencing the
for the uncertainty associated with resource availability, which work of Federici et al. who surveyed a population of US
we optimize for recommended QOL decisions to inform base Army base camp subject matter experts and employed conjoint
camp design doctrine. [20], [25] analysis to establish relative ”part-worth” values for each QOL
decision. [5] Our methodology was also informed by previous
C. Summary of Contributions work concerning the impact of resource constraints on QOL
Our investigation provides three contributions. First we by Augustyn et al. [2]
formulate a specific methodology for improving base camp The formulation of the BCMIP-CC provides the capability
QOL using the BCMIP. To validate the performance of the to optimize QOL while accounting for resource uncertainty.
BCMIP we compare our recommended QOL decisions to To develop this formulation we employed a chance constraint
current US Army QOL TTPs. Next we investigate how QOL optimization model influenced by the work of Ruszczynski.
is affected by changes in the BCMIP parameters including [25]
resource availability, the marginal usage of resources with each
B. Base Camp Mixed Integer Program (BCMIP)
QOL decision, and the contribution of each QOL decision
towards soldier mission effectiveness. Finally, we provide a Our initial methodology employed a mixed integer program
methodology for the optimization of base camp QOL under to provide optimized QOL decisions on US Army base camps
uncertain resource constraints by developing the BCMIP-CC. under the deterministic constraints of fuel, water, waste water,
The BCMIP-CC provides decision makers with parameters and solid waste. We refer to this initial formulation as the
that can be used to optimize base camp design given terrain Base Camp Mixed Integer Program (BCMIP). Current US
restrictions and the mission. By refining and employing the Army base camp design produces a lower QOL score when
BCMIP-CC methodology we believe US Army decision mak- compared to the decisions recommended by the BCMIP. The
ers have the tools to develop base camp design doctrine. QOL score is produced by an objective function and used
to compare the QOL provided to soldiers given varying QOL
II. T ECHNICAL A PPROACH decisions. [5] With a fixed amount of resources, our optimized
US Army decision makers design individual base camps, QOL decisions achieve a score of 16.5 compared with the
and hence influence QOL based on personal experience and current baseline score of 13.7 - a 20% increase in QOL with no
current TTPs. [2], [18] We build upon this approach by additional resource consumption. The baseline QOL decisions
developing the BCMIP, which determines optimized QOL that are currently employed on base camps and those decisions
decisions that improve soldier mission effectiveness under recommended by the BCMIP are provided in Table I.
constraints on resource availability. This methodology lever- QOL Current Standards Optimized Decisions
ages both the qualitative knowledge of a population of base Decision with 100% resources with 100% resources
camp subject matter experts and mathematical optimization. Shower Hot 10 minute Hot 5 minute
We accomplish this by first using a QOL function developed Laundry Bags 2 per week 1 per week
by Federici et al. at Natick Labs. [5] Next, we calculate the Latrines Flush Flush
Breakfast MRE UGR-A
marginal utilization of each QOL decision for each limited
Lunch MRE MRE
resource, establish historical resource availability levels, and Dinner UGR-A UGR-A
formulate the BCMIP to determine optimized QOL decisions Refrigerated
for base camps. Finally, we formulate chance constraints in the Beverages Not Available Available
BCMIP-CC to find the best QOL decisions given uncertainty Ice Available Not Available Not Available
associated with contingency operations. Enhanced Food Not Available Available
Gym Cardio Cardio and Weight
A. Related Work MWR Basic Improved
Mail Facility Available Available
Literature from both civilian and military investigations HVAC High High
informed our understanding of soldier QOL and the resource QOL Score 13.7439 16.572
constraints encountered on US Army base camps in contin-
gency operations. Medical journal articles were critical to our TABLE I: Quality of Life Decision Comparison[6], [9], [12],
evaluation of the QOL model developed at Natick Labs SLB- [13]
STO-D, including work by Spitzer et al. and Ramsey et al.
[5], [22] Military QOL studies were informative given our To evaluate the sensitivity of the BCMIP results given
application of a qualitative function to a military population. uncertainty in the model parameters, we performed numerical
sensitivity analysis. In this sensitivity analysis we perturbed all QOL and resulting soldier mission effectiveness on base
the parameter values of the BCMIP individually, iteratively re- camps is constrained by a set of fundamental resources that
optimizing the BCMIP with each perturbation. We compared impose an enormous burden on contingency operations in mis-
the results of each re-optimization to previous iterations. We sion capabilities and financial costs. This set of constraining
determined that QOL on US Army base camps is most resources, R, is provided in Table III.
sensitive to the fuel and water constraints. The details of the
BCMIP formulation and sensitivity analysis are available in Members of Set R: Limited Resources
j Rj Resources Unit
Kuiper. [17] gallons
1 Fuel Diesel Fuel
This result aligned with practical research in the design and soldier daily
gallons
2 Water Water
sustainment of US Army base camps in contingency opera- soldier daily
gallons
tions where the supply of fuel and water acutely impacted the 3 Waste Water Waste Water soldier daily
pounds
QOL. [20] With these results we developed a new formulation 4 Solid Waste Solid Waste soldier daily

using chance constraints (BCMIP-CC) in order to account for

the uncertainty associated with fuel and water supply. TABLE III: R ≡ set of limited resources available on a base
In the next section we describe the formulation of BCMIP- camp
CC, which includes chance constraints on resource availability.
C. BCMIP-CC Formulation To account for the uncertainty associated with resource
1) Sets: The set of QOL decisions established by decision supply, many decision makers naively under-estimate the
makers designing base camps is denoted by S. These QOL amount of resources available when designing base camps.
decisions are the services on base camps that have the most For example a decision maker may design a base camp
impact on solider mission effectiveness in contingency opera- assuming 2 gallons of fuel available per day per soldier, while
tions and are listed in Table II. [17] actually expecting to receive 3 gallons of fuel per day per
soldier. Naively planning for a lower amount of water or
Members of Set S: Quality of Life Decisions fuel when designing base camp QOL decisions is designed
i Si Description
1 Cold 5 minute Shower 10◦ temp change for 5 minutes to account for the possibility of missed resupply or inefficient
2 Warm 5 min Shower 20◦ temp change for 5 min service. While this method may provide an adequate surplus
3 Hot 5 min Shower 30◦ temp change for 5 min of resource supply, it does not give decision makers the
4 Cold 10 min Shower 10◦ temp change for 10 min tools to make optimized base camp design decisions based
5 Warm 10 min Shower 20◦ temp change for 10 min
6 Hot 10 min Shower 30◦ temp change for 10 min upon operational requirements. A more rigorous methodology
7 Cold 15 min Shower 10◦ temp change for 15 min that mitigates uncertainty is to employ a relative frequency
8 Warm 15 min Shower 20◦ temp change for 15 min distribution of resource availability, and to optimize QOL
9 Hot 15 min Shower 30◦ temp change for 15 min
10 Number of Laundry Bags Number of Laundry Bags Daily
under the assumption of this distribution with some level of
11 Burn Out Toilets Burn Disposal Latrines certainty. [23].
12 Chemical Toilets Chemically Treated Latrines We establish the set Ω of resource availability scenarios
13 Flush Toilets Latrines with Full Plumbing
14 Breakfast MRE Warm Meal for Breakfast
for the base camp and vary the availability of water and
15 Breakfast UGR-E Improved Combat Ration for Breakfast fuel according to the probabilistic event of missed resource
16 Breakfast UGR-A Basic Combat Ration for Breakfast deliveries. For example, referencing Table IV for Scenario 10
17 Lunch MRE Warm Meal for Lunch we see a joint event has occurred where 2 days’ fuel and 1
18 Lunch UGR-E Improved Combat Ration for Lunch
19 Lunch UGR-A Basic Combat Ration for Lunch day’s water delivery have been missed. As a result, there will
20 Dinner MRE Warm Meal for Dinner be a 75% decrease in available fuel and 50% decrease water
21 Dinner UGR-E Improved Combat Ration for Dinner on the base camp.
22 Dinner UGR-A Basic Combat Ration for Dinner
Refrigerated
23 Refrigerated Beverage Beverages Available Members of Set Ω
24 Ice Available Bagged Ice Available at all Times Missed
deliveries No days 1 day 2 days 3 days
Additional Food
25 Enhanced Food Available between Meals by resource missed fuel missed fuel missed fuel missed fuel
Fitness Center with No days
26 Cardio Gym Cardiovascular Equipment missed water Scenario 1 Scenario 5 Scenario 9 Scenario 13
27 Weight Gym Fitness Center with Weight Equipment 1 day
28 Basic MWR Small MWR with 2 TVs missed water Scenario 2 Scenario 6 Scenario 10 Scenario 14
29 Improved MWR Large MWR with 6 TVs 2 days
30 Mail Facility Distribute and Hold Mail
missed water Scenario 3 Scenario 7 Scenario 11 Scenario 15
31 Medium HVAC Medium Level of HVAC
3 days
32 High HVAC High Level of HVAC
33 Laundry Indicator Indicator Variable for Laundry missed water Scenario 4 Scenario 8 Scenario 12 Scenario 16
34 Hot Meal Indicator Indicator Variable for Hot Meals
TABLE IV: Ω ≡ set of scenarios describing the availability
TABLE II: S ≡ set of camp quality of life decisions of resources
 Each set member ω ∈ Ω represents a scenario that specifi-
cally describes the resources available to the base camp, which Resource Availability Distribution :
is determined by the joint probabilistic event of whether a
d(1 − d)x−1 d(1 − d)y−1
planned fuel and / or water delivery does or does not occur pj,ω = P(X = x, Y = y) = ( )( ),
daily. We reduced the availability of fuel and / or water by 1 − (1 − d)n 1 − (1 − d)n
−50% in the event that a delivery to the base camp is missed. j, ω ∈ R, Ω; and n = 4 days
2) Parameters: Parameters provide the numerical specifica- (2)
tion for the BCMIP-CC. These values are indexed by the three The final parameters of the BCMIP-CC, are the resource
sets we previously established: QOL decisions S, resources resupply uncertainty value d ∈ [0, 1], used in Equation 2, and
R, and resource availability scenarios Ω. We introduce six the likelihood of infeasibility of the results or risk associated
parameters for the BCMIP-CC: QOL objective function co- with the model, represented by j ∈ [0, 1], where j ∈ R. For
efficients, resource availability values, marginal use values, the model we evaluated we assumed a constant j =  ∀j ∈ R.
a discrete probability distribution for resource availability, 3) Decision Variables and Objective Function: The deci-
resupply uncertainty value, and risk. We provide a description sion variable xi is the QOL decision, xi ∈ {0, 1}, i ∈ S,
of these parameters; exact values may be referenced in Base where xi = 1 if decision i is selected and xi = 0 otherwise.
Camp Standardization and Improvement by Kuiper. [17] The decision variable x10 is the exception to this definition,
The QOL objective function coefficients, qi ∈ R+ , where where x10 ∈ Z, or the set of integers.
i ∈ S, describe how much each QOL decision contributes Our decision variable zj,ω is the indicator variable for the
towards a soldier’s QOL. These values were developed by robustness of the BCMIP-CC to the availability of resource
surveying a population of base camp subject matter experts and j ∈ R for scenario ω ∈ Ω, where zj,ω ∈ {0, 1}. zj,ω = 0
employing conjoint analysis to develop part-worths of each if the BCMIP-CC is robust to scenario ω, or zj,ω = 1 if the
QOL decision. [19] These coefficients values were taken from BCMIP-CC is not robust to scenario ω.
an analysis executed by Natick Labs SLB-STO-D. [5] The BCMIP-CC assumes a linear relationship between de-
The parameter bj,ω describes the availability of each re- cision variables. With this relationship, and referencing Natick
source, j ∈ R, given a specific resource availability scenario, Labs SLB-STO-D research, our model uses a linear objective
ω ∈ Ω, where bj,ω ∈ R+ . In Equation 1 we provide the function employing the QOL coefficients qi , to be multiplied
relationship used to calculate resource availability. by each QOL decision: x1 q1 + x2 q2 + . . . + xn qn . [5]
We refer to the value generated by this function as the QOL
Resource Availability Calculation : score. This is the function we optimize in order to determine
bj,ω = (bj, Scenario 1 ) × (lmj,ω ), where: improved QOL decisions.
l = loss factor constant of .5
Total Quality of Life (QOL Score)
mj,ω = number of missed resupply for resource j X
and scenario ω; where mj,ω ∈ {0, 1, 2, 3}† qi xi (3)
† i∈S
Reference T able IV for scenario specification of mj,ω
(1) In order to determine the optimized or Recommended QOL
The marginal use parameter quantifies how much each QOL Decisions for doctrinal base camp design we find the Argu-
decision costs in terms of each resource. For example if ment Maximum (arg max) of the objective function, Equation
decision makers design base camps with the QOL decision 3, of the BCMIP-CC.
of 10 minute cold showers, then a specific amount of fuel 4) Constraints: Analyzing the relationship between each
is consumed to pump and heat water, water is consumed to QOL decision and limited resources is critical to improving
provide the shower, and waste water is produced as a result of soldiers’ mission effectiveness while executing contingency
the shower. The marginal use values for each QOL decision operations. We define linear constraint relationships, quanti-
were calculated referencing consumption models developed fying the amount that each quality of life decision variable,
by Natick Labs SLB-STO-D. We represent these values with xi , consumes of each resource availability scenario, bj,ω ,
ui,j ∈ R+ where i, j ∈ S, R. measured per person daily.
The parameter pj,ω ∈ [0, 1], j, ω ∈ R, Ω specifies The marginal use values for each resource, ui,j , provides
the discrete probability distribution of resource availability. the multiplicative link between QOL decisions and resource
These probabilities quantify the likelihood of each resource availability. The resource availability chance constraint rela-
availability scenario. Our scenarios, indexed by set Ω, only tionships ensure that the QOL decisions recommended by the
describe the possibility of reduced fuel and water levels. BCMIP-CC do not consume more resources then available
We employed a joint truncated geometric distribution, where on the camp. The amount of resources available is provided
mf uel,ω = X ∈ {0, 1, 2, 3} is the random variable of the by the discrete distribution of resource availability scenarios.
number of water deliveries missed and the random variable A generalized constraint inequality considering all resource
mwater,ω = Y ∈ {0, 1, 2, 3} is the number of fuel deliveries availability scenarios, bj,ω and all potential QOL decisions, xi
missed. is provided in Equation 4.
 III. R ESULTS AND E XPERIMENTATION
Resource Chance Constraints :
A. Sensitivity analysis of Deterministic Model
X
ui,j xi + zj,ω bj,ω ≤ bj,ω ∀ j, ω ∈ R, Ω (4) Prior to formulating the BCMIP-CC we evaluated the
i∈S assumptions of the deterministic BCMIP. To conduct this
evaluation we established a baseline of current QOL TTPs
With the BCMIP-CC formulation the resource availability
decisions specifying how the US Army currently designs base
levels are not fixed; instead a distribution of scenarios that
camps to evaluate the performance of the BCMIP. To build
provide varying levels of resources are used to constrain the
this QOL baseline we referenced several regional command
BCMIP-CC. The probability constraint, shown in Equation
Standard Operating Procedure (SOP) documents and the cur-
5 provides additional resources to the BCMIP-CC, loosening
rent Base Camp US Army Field Manual (FM 3-37.20). [13],
the constraints associated with Equation 4 as the BCMIP-CC
[12], [6] For an initial performance evaluation, we optimized
model optimizes over the binary decision variable zj,ω . The
the BCMIP and compared the resulting recommended QOL
additional resources provided to the model are controlled by
decisions to the QOL score achieved by current QOL TTPs
the risk parameter . With the relationship between Equations
decisions (measured by the objective function developed by the
4 and 5 we see that  controls what resource availability
Natick Labs SLB-STO-D). Table I provides this comparison of
scenarios the BCMIP-CC does not account for.
the optimized QOL decisions determined by the BCMIP (the
arg max of the BCMIP) and current QOL TTPs, including
Probability Constraint : QOL scores for performance evaluation.
X With the baseline QOL TTPs decisions established, we
pj,ω zj,ω ≤ j ∀ j ∈ R (5) performed numerical sensitivity analysis on the BCMIP to
ω∈Ω evaluate the assumptions associated the parameters of the
model. Numerical sensitivity analysis was performed by per-
In addition to these two primary constraints, there are a turbing each parameter individually by ±1% and iteratively
number of logical constraints that are formulated to enforce resolving the BCMIP. Each iteration of optimized solutions
the physical structure of base camps. These constraints in- were then compared to the initial results of the BCMIP and
clude: fixed cost constraints associated with generator imple- the baseline QOL TTPs decisions. We first evaluated the
mentation, policy restrictions on base camps, and decision sensitivity of the objective function coefficients parameters qi
variable restrictions. The full formulation of the BCMIP-CC and marginal use coefficients ui,j . We found that the model
is described in detail by Kuiper and provided in Table V. [17] was insensitive to changes in both these parameters; the details
of this analysis can be found in Kuiper. [17]
MAXIMIZE:
Description Equation Next, we evaluated the sensitivity of the BCMIP to resource
Total QOL P availability, bω,j where ω = {Scenario 1} ∀ ω ∈ Ω for the
Score i∈S qi xi BCMIP, by iteratively decreasing all resources by 1% and
SUBJECT TO:
Resource P
resolving the BCMIP with each iteration. Once again we
Chance Constraint i∈S ui,j xi + zj,ω bj,ω ≤ bj,ω ∀ j, ω ∈ R, Ω compared each iteration to previous optimized solutions and
Probability
P the baseline. We found that fuel, water, waste water, and
Constraint ω∈Ω pj,ω zj,ω ≤ j ∀ j ∈ R
Shower solid waste could all be reduced by 29% uniformly while
Constraint x1 + x2 + x3 + x4 + x5 + x6 + x7 + x8 + x9 ≤ 1 maintaining an equivalent QOL when compared to the score
Breakfast Choice x14 + x15 + x16 = 1 achieved by the baseline. In Figure 1 we provide the results
Lunch Choice x17 + x18 + x19 = 1
Diner Choice x20 + x21 + x22 = 1 of this analysis.
Latrine Choice x11 + x12 + x13 = 1
MWR Choice x27 + x28 = 1
HVAC Choice x31 + x32 = 1 18 Quality of Life Sensitivity with Uniform Resource Change
Laundry Fixed 7x33 ≥ x10
Laundry Bag Limit x10 ≤ 7 16
Meal Fixed Cost x16 + x19 + x22 ≥ x34
Quality of Life Value

Meal Fixed Cost x16 + x19 + x22 ≤ 3x34 14

Binary Decision
Variables x1−9 ∈ {0, 1} 12
Integer Decision
10
Variables x10 ∈ Z1
Binary Decision US Army Standard
8 Optimized Quality of Life
Variables x11−34 ∈ {0, 1} Equvalent QoL with 29% Reduction
Chance Constraint 610 0 10 20 30 40 50 60 70 80
Indicator zj,ω ∈ {0, 1} ∀ j, ω ∈ R, Ω Percent Decrease in Resources Available

TABLE V: Complete Formulation of the BCMIP-CC Fig. 1: Resource Availability Sensitivity Analysis
 The results of the resource availability sensitivity analysis they are willing to accept when considering the event that the
confirmed the utility of employing the BCMIP to determine camp does not receive planned resource resupply. The BCMIP-
QOL decisions when designing base camps. This analysis also CC provides decision makers with the optimization power of
indicated that base camp QOL is sensitive to the resource the BCMIP along with two critical tuning parameters. The
availability parameter, bω,j . After investigating the sensitivity resource resupply uncertainty value d allows decision makers
of the BCMIP for each resource j ∈ R we determined that to adjust the resource availability uncertainty associated with
base camp QOL and soldier mission effectiveness is most a camp and is an exogenous variable relative to the decision
impacted by fuel and water availability levels. Figures 2 and maker. The risk tolerance parameter  adjusts the buffer
3 provide QOL scores achieved by the BCMIP as fuel and (additional resource supply beyond daily demand) of resources
water resource availability levels are reduced. Each iteration of kept on a base camp, serving as an endogenous variable for the
perturbation that results in a QOL score change indicates that decision maker. Choosing parameters based upon the mission
the recommended QOL decisions have changed. We’ve labeled and surrounding terrain of a base camp and then optimizing
each of these decision points with an integer for reference. the BCMIP-CC allows decision makers to guard against the
event of reduced resources when designing camps.
17 Quality of Life Fuel Sensitivity The uncertainty associated with the resupply of a base camp
1
Fuel Data 3 2 as specified in Equation 2 is quantified by the parameter d.
16 US Army Standard 6 5
4
Decision Points 8 7
This parameter represents the ”success” parameter of Bernoulli
Quality of Life Value

15 10 9 trial, where ”success” is the successful resupply of a limited

11
12
13 resource daily to the base camp and ”failure” is the missed
14 14
1615 resupply of a limited resource daily. The decision maker may
13 17
adjust d smaller (closer to 0) as the terrain surrounding a camp
1918
12 2120 becomes more difficult to travel through to deliver resources.
110.5
Terrain that is difficult to travel through, or restricted terrain,
1.0 1.5 2.0 2.5 3.0 3.5 naturally causes resource resupply to become more uncertain
Fuel Available in Gallons
leading to a decreased d value. As expected, decreasing the
Fig. 2: Fuel Resource Availability Sensitivity Analysis parameter d closer to 0 will decrease the QOL score of a base
camp because missing the resupply of a resource becomes
more probable, reducing the level of available resources across
Quality of Life Water Sensitivity scenarios ω ∈ Ω.
17.0
Water Data 1
16.5 US Army Standard 2 The parameter  is a measure of a camp’s guard against
3
16.0 Decision Points uncertainty quantifying a decision maker’s risk tolerance for
Quality of Life Value

4
15.5 5 a base camp. This parameter ensures that a camp maintains
6
15.0 7 a resource buffer amount, which is increased for a low risk
8
14.5 tolerance camp, and decreased for a high risk tolerance camp.
10 9
14.0 A buffer is a reserve of limited resources, consumed when
11 a resupply is missed. Risk tolerance for a decision maker
13.5 1312
13.00 when designing a base camp is dictated by the mission,
5 10 15 20 25
Water Available in Gallons for example a complex camp with mission essential aircraft
providing international transportation would have much lower
Fig. 3: Water Resource Availability Sensitivity Analysis risk tolerance than a small simple camp of rifleman.
If all other parameters are kept constant, lowering the risk
These results align with previous research in the design
tolerance for a base camp will result in a reduced QOL score.
and sustainment of US Army base camps in contingency
This is a result of an increased resource buffer, which reduces
operations where the supply of fuel and water acutely im-
the amount of resources available to provide QOL services
pacted soldier QOL and in turn mission effectiveness. [3],
on the camp. A larger buffer provides a stock of resources
[20] This sensitivity analysis of resource availability, coupled
available to maintain QOL services in the event of missed
with previous research concerning the uncertainty associated
resource resupply. The BCMIP-CC allows decision makers to
with contingency operations, motivated the development of the
quantify the risk they are accepting when designing a camp by
BCMIP-CC.
sacrificing QOL in order to guard against uncertainty. When
B. Analysis of Chance Constraint with Varying Parameters a decision maker lowers the risk tolerance of a base camp
design, the BCMIP-CC increases the buffer of resources avail-
US Army decision makers must address two fundamental
able on the camp to guard against missed resource resupply.
concerns when designing a base camp: 1) how to deal with the
uncertainty associated with missing the resupply of resources, In Table VI we provide the QOL decisions for the optimized
often dictated by terrain restrictions, and 2) the level of risk BCMIP-CC for two examples of base camps.
 Large Complex Small Simple
when designing base camps. When compared to relying upon
Base Camp Base Camp
Decision j = .05, d = .9 j = .25, d = .5
individual anecdotal experience, the BCMIP-CC is a signif-
Shower Hot 5 minute None icantly more effective method for establishing base camp
Laundry Bags None None design doctrine. The BCMIP-CC parameterizes the uncertainty
Latrines Chemical Toilets Chemical Toilets and risk associated with the QOL design decisions of a
Breakfast UGR-A UGR-A base camp, allowing decision makers to deliberately construct
Lunch UGR-A UGR-A efficient camps, while maintaining QOL.
Dinner UGR-A UGR-A
Refrigerated V. F UTURE W ORK
Beverages Not Available Available
Ice Available Available Not Available A priority of our future work is to explore the impact of
Enhanced Food Available Available changes to the parameters  and d on the results of the BCMIP-
Gym Cardio Cardio and Weight CC. Furthermore, we would like to investigate how to estimate
MWR Basic Improved numerical values for  and d specific to mission sets and ter-
Mail Facility Not Available Available
rains restrictions associated with actual base camps. We would
HVAC High None
Score 13.946 12.013 like to determine the correlation between measurable mission
effectiveness of soldiers and the QOL score determined by the
TABLE VI: Comparison of Chance Constraint Analysis BCMIP-CC. This would return the quantitative QOL score
back to a qualitative metric with practical meaning to US
The camps described in Table VI have distinctly different Army decision makers.
mission sets and resource supply uncertainty. The Large Com- Leveraging the final formulation and results of the BCMIP-
plex Base Camp models a camp where resource availability CC, we believe the US Army can define doctrinal standards
occurs with little interruption, d = .9, and a low tolerance for for QOL decisions. US Army decision makers should review
risk,  = .05. The low tolerance for risk models a complex this investigation, with previous research, to develop doctrinal
camp with essential assets or where soldiers rest and reset base camp design standards with regard to QOL. Establishing
equipment. Additionally, a missed resupply on larger more well defined doctrine employing the methodology developed
complex camps could prove catastrophic given the construc- by this investigation, will improve soldier QOL, mission
tion of more complex QOL services, further necessitating a effectiveness, and reduce the resources required to execute
low risk parameter. global contingency operations.
The second camp models a Small Simple Base Camp,
ACKNOWLEDGMENT
with a much higher likelihood of missed resource resupply
and a higher risk tolerance. This parameterization accurately We would like to thank the continued support of this re-
models a forward camp with basic QOL services, where basic search by Draper, Natick Labs, the Department of Mathematics
services are easily adjusted in the event of a missed resupply at the United States Military Academy, the Army Research
when compared to larger complex camps. Laboratory, and the John A. Paulson School of Engineering
and Applied Sciences at Harvard University. These organiza-
IV. C ONCLUSION tions have been generous with time and resources throughout
The QOL decisions employed on base camps during our investigation.
Operation Endure and Iraqi Freedom consumed unsustainable
amounts of resources. [3], [18] While consuming large R EFERENCES
amounts of resources these camps delivered relatively little [1] H. G. Anderson, and K. Kinnevan, ”Contingency Base Camp Operations
return in QOL as shown by our analysis of current US Army and Management: Staffing and Organization,” ERDC/CERL TR-13-18,
QOL TTPs. Subsequently, current base camp design and QOL Sep. 2013.
[2] J. Augustyn, P. Benasutti, S. Bolduc, D. Darkow, J. Federici, R. Haddad,
decisions are challenged by the results of the BCMIP-CC. J. Howe, P. Mccarty, and J. Mahoney, ”Technology-Enabled Capa-
While the results of the deterministic BCMIP demonstrated bility Demonstration 4A Sustainability and Logistics-Basing Baseline
improvement over current practices, our follow-on sensitivity Wargame,” US Army Natick Soldier Center, 2012.
[3] COL H. G Anderson, ”Sustainable Base Camp Contingency Operations,”
analysis showed that our initial recommended QOL decisions USACE ERDC-CERL, Jun. 2010.
were sensitive to changes in fuel and water, leading to our [4] C. Correia, A. Horner, J. Mclaughlin, and D. Stewardson, ”The Chal-
recommendation that base camp design account for the lenges Associated with Accounting for the Armys Force Provider System
when Deployed in Support of Military Operations,” Naval Postgraduate
possibility of lower fuel and water supply. The BCMIP-CC School, Sep. 2008.
addressed this uncertainty and we provide two examples [5] J. Federici, L. Lesher, J. Augustyn, and J. Howe, ”Technology-Enabled
of optimized QOL decisions, with uncertainty and risk Capability Demonstration 4A Sustainability and Logistics-Basing: Initial
Quality of Life and Soldier Readiness User Assessment,” U.S. Army
parameters d and  in Table V I. Natick Soldier Research, Development and Engineering Center, Oct.
2014.
The BCMIP-CC is the most important contribution of this [6] Headquarters, Department of the Army, ”Base Camps,” Army Techniques
Publication (ATP) 3-37.10.(Final Draft) Washington, DC: HQDA, 2011.
investigation; employing a distribution of resource availability [7] Headquarters, Department of the Army, ”Brigade Combat Team,” Field
is critical to providing effective and efficient QOL decisions Manual (FM) 3-90.6. Washington, DC: HQDA, 2010.
[8] Headquarters, Department of the Army, ”Counterinsurgency,” Field
Manual (FM) 3-24. Washington, DC: HQDA, 2006.
[9] Headquarters, Department of the Army, ”Force Provider Operations,”
Army Techniques Publication (ATP) 4-45 Washington, DC: HQDA, Nov.
2014.
[10] Headquarters, Department of the Army, ”Intelligence Preparation of the
Battlefield,” Field Manual 34-130, Jul. 8, 1994.
[11] Headquarters, Department of the Army, ”Terms and Military Symbols,”
ADRP 1-02, Dec. 2015.
[12] Headquarters, United States Army Europe and Seventh Army, ”US
Army Pamphlet 420-100: Standards for Forward Operating Sites,” United
States Army Installation Management Command Europe Region, Apr. 21,
2008.
[13] Headquarters, United States Central Command, ”Construction and
Base Camp Development in the USCENTCOM Area of Responsibility,”
Regulation Number 415-1, 2009.
[14] G. Jandzinski, ”Identifying Situational Factors Contributing to
Combat Performance During Desert Shield and Desert Storm,”
AFIT/GAL/LAR/95S-5, 1995.
[15] Joint Chiefs of Staff, ”Joint Operations,” US Department of Defense,
Aug. 11, 2011.
[16] D. A. Krooks, L. A. Whalley, and H. G. Anderson, ”Contingency Bases
and the Problem of Sociocultural Context,” ERDC/CERL TN-12-2, 2012.
[17] P. K. Kuiper, ”Base Camp Quality of Life Standardization and Im-
provement,” Submitted in Partial Fulfillment of a Masters of Engineering,
Harvard University, 2016.
[18] MAJ S. Marsh Read, ”Base Camp Life Cylce Management: Focusing on
the Critical Elements,” School of Advanced Military Studies, 2011-2012.
[19] N. Malhotra, and D. Brooks, Marketing Research: An Applied Approach.
Milan, Italy: Prentice Hall, 2007.
[20] Noblis Corporation, ”Sustainable Forward Operating Bases,” Strategic
Environmental Research and Development Program, May 21, 2010.
[21] E. Peltz, M. Robbins, K. Girardini, R. Eden, J. Halliday, and J. Angers,
”Sustainment of Army Forces in Operation Iraqi Freedom,” RAND
Corporation, 2005.
[22] S. Ramsey, R. Etzioni, A. Troxel, and N. Urban, ”Optimizing Sampling
Strategies for Estimating Quality-adjusted Life Years,” Medical decision
making, Oct. 1997.
[23] D. Reich, ”A Linear Programming Approach for Linear Programs with
Probabilistic Constraints,” European Journal for Operations Research,
230 (2013) 487494, Jun. 2012.
[24] R. V. Robinson, ”Combat Service Support Reach: A Risky Support
Strategy for the Interim Brigade Combat Team,” US Army War College,
USAWC Strategy Research Project, 2002.
[25] A. Ruszczynski, ”Probabilistic Programs with Discrete Distributions
and Precedence Constrained Knapsack Polyhedra,” Rutgers Center for
Operations Research, 2003.
[26] W. Spitzer, A. Dobson, J. Hall, E. Chesterman, J. Levi, R. Shepard, R.
Battista, and B. Catchlove, ”Measuring the Quality of Life of Cancer
Patients,” Royal North Shore Hospital of Sydney, The Commonwealth
Institute of Health of Australia at Sydney University, The University of
Newcastle and the McGill Cancer Centre. Montreal, Canada, Nov. 1980.
[27] Submitted by: John M. McHugh and General Raymond T. Odierno, ”The
United States Army Posture Statement,” The United States Senate and
the House of Representatives, Apr. 2014.
[28] ”Transforming Wartime Contracting: Controlling Costs, Reducing
Risks,” Commission on Wartime Contracting in Iraq and Afghanistan,
Aug. 2012.
[29] Untied States Army Natick Laboratories, ”TECD 4a Baseline Summary,”
US Army NSRDEC, May 8, 2014.
[30] United States Government, ”United States Code, Supplement 5, Title
10 - ARMED FORCES,” Bills and Statutes, 2006 Edition.
[31] USAREUR DCG, ”Base Camp Facility Standards: The Red Book,”
Contingency Operations, Nov. 2001.
[32] D. Vergun, ( 2014, Sep. 10). McMaster busts myths of future warfare
[Online]. Available: [Link]
[33] G. Wilcove, D. Wolosin, and M. Schwerin, ”Development of a New
Quality of Life Model Using Structural Equation Modeling,” Navy
Personnel Research, Studies, and Technology Department, Sep. 2002.
[34] W. Winston, and Venkataramanan, Introduction to Mathematical Pro-
gramming. Pacific Grove: Thompson, Brooks, Cole, 2003.