See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/384009028

Space Logistics Modeling and Optimization: Review of the State of the Art

Article in Journal of Spacecraft and Rockets · September 2024

DOI: 10.2514/1.A35982

CITATIONS READS
0 12

1 author:

Koki Ho
Georgia Institute of Technology
141 PUBLICATIONS 1,201 CITATIONS

SEE PROFILE

All content following this page was uploaded by Koki Ho on 14 September 2024.

The user has requested enhancement of the downloaded file.

 Space Logistics Modeling and Optimization:
Review of the State of the Art∗

Koki Ho†
Georgia Institute of Technology, Atlanta, GA, 30309

I. Introduction
The US Space Force’s doctrine document published in 2020, SpacePower: Doctrine for Space Forces listed five

core competencies that the US Space Forces need to perform, one of which was “Space Mobility and Logistics” [1]. In

response to that, there has been a rapidly growing interest in developing technologies to enable in-space operations via

in-space infrastructures and in-space refueling, servicing, assembly, and manufacturing (ISAM). Space logistics has also

been a topic of interest to NASA and other space agencies as an approach to designing sustainable space exploration

campaigns. Additionally, there is a growing need for space logistics to effectively operate the mega-scale constellations

in Earth orbits for communication, remote sensing, and related missions.

Enabling space mobility and logistics capabilities requires a new way to view space missions that differs from

conventional astrodynamics. For example, with fuel depots in space, the optimal path to the destination in terms of the

lifecycle cost is not necessarily a fuel-optimal trajectory; rather, a path stopping by a fuel depot and being refueled

before heading to the destination may be preferred even when it requires additional fuel. In addition, with ISAM

capabilities, we need additional analysis capabilities to analyze and optimize the sizes of the fuel/spare depots and

their inventory/sparing policies with astrodynamics in mind. This analysis requires logistics-driven modeling and

optimization techniques coupled with astrodynamics. This intersection between these areas leads to an emerging field,

Space Logistics modeling and optimization.

Space Logistics is formally defined as “the theory and practice of driving space system design for operability and

supportability, and of managing the flow of materiel, services, and information needed throughout a space system

lifecycle [2]” according to the AIAA Space Logistics Technical Committee. Space Logistics is not a new research topic;

the literature on space logistics has existed since the beginning of the space exploration era [3–5]. According to the

AIAA’s meeting papers archive, the 1st AIAA/SOLE Space Logistics Symposium was held in 1987 [6], and it already

had papers that covered various topics that are still relevant today such as in-space servicing [7, 8] and even using

artificial intelligence (AI) for space logistics [9]. Although the applications of interest back then do not necessarily

match with our current ones, various key concepts were developed decades ago. In the early 2000s, in response to

NASA’s Constellation Program, the idea of logistics network modeling was applied to human space mission design. In
∗ The previous version of this paper was presented at the AIAA SciTech Forum in Orlando, FL, on Jan 8-12 (AIAA 2024-1275).
† Dutton-Ducoffe Professor, Associate Professor, Daniel Guggenheim School of Aerospace Engineering, Chair of the AIAA Space Logistics
Technical Committee, AIAA Senior Member
addition, more logistics-driven techniques for probabilistic modeling and inventory control have been applied to satellite

servicing and mega-scale constellations. A recently published book, The Planning and Execution of Human Missions to

the Moon and Mars [10], has two chapters that focus on space logistics and mathematical modeling for space mission

design [11, 12], which can be referenced for more details.

The goal of this paper is to categorize the state-of-the-art studies in Space Logistics modeling and optimization

in two ways: (1) by application questions that are addressed; and (2) by logistics-driven methods that are used in the

studies. The goal of the first categorization is to help the practitioners determine what research is out there that can

support their applications. The goal of the second categorization is to systematically map the existing literature to each

logistics research subfield, and thus help the researchers to understand the state of the art and identify the under-explored

and promising future research directions.

II. Categorization of State of the Art by Applications

In this section, we first categorize the state-of-the-art literature by applications to aid the practitioners in understanding

the existing research that can answer their application questions. We will introduce the three major applications in Space

Logistics, and then map the state of the art in the literature to each research question.

Note that this is by no means a comprehensive list of all space logistics applications. This paper specifically focuses

on the applications that involve orbital mechanics considerations; other space logistics applications not reviewed in this

paper include supportability and spare parts management within a space station [13–15], surface vehicle routing and

logistics [16–23], among others.

A. Application 1 (A1): In-space Servicing, Assembly, and Manufacturing (ISAM) for Satellites

Although In-space Servicing, Assembly, and Manufacturing (ISAM) has attracted a lot of attention in recent years,

the concepts themselves have been explored for a long time. When an operational satellite experiences a component

failure or runs out of its fuel, it could be economical to repair and refuel the satellite on-orbit than launch a new

one from the ground especially when the satellite is in a high-altitude orbit. One of the earlier successful examples

of on-orbit servicing was the repair mission for the Hubble Space Telescope, although crewed repair missions have

become difficult after the retirement of the Space Shuttle. In the 2000s, multiple studies introduced the modeling and

simulation foundations for robotic/autonomous on-orbit servicing missions [24–26] and examined the feasibility and

economic analysis of various ISAM concepts [27, 28]. More recently, robotic on-orbit servicing has also been explored

by DARPA in its Robotic Servicing of Geosynchronous Satellites (RSGS) program [29] as well as by industries such as

Northrop Grumman [30]. In addition, in-space manufacturing has also been explored by DARPA’s Novel Orbital Moon

Manufacturing, Materials, and Mass Efficient Design (NOM4D) Program [31] as well as other industry members such

as Redwire.

2
 In parallel to the hardware technologies for robotic ISAM, there is a significant need for logistics-driven approaches

in the context of ISAM. The key questions that this paper will focus on include:

• A1Q1: how to analyze the performance of an ISAM architecture;

• A1Q2: how to tactically plan and schedule the ISAM operations to satisfy the uncertain demands of the customer

satellites under resource constraints; and

• A1Q3: how to strategically architect the ISAM infrastructure elements such as depots and vehicles.

These questions can be answered through logistics-driven research as reviewed later.

B. Application 2 (A2): Multi-Mission Space Exploration Campaigns

Besides ISAM for satellites in Earth orbits, we are also interested in how space exploration campaigns can leverage

in-space logistics infrastructure elements. The concept of an in-space propellant depot for space exploration has been

studied extensively (e.g., [32–35]). In addition, the recent development of technologies for in-situ resource utilization

(ISRU) on the Moon and Mars has become a game changer [36–39]. For example, the successful oxygen generation

on Mars by the Mars Oxygen In-Situ Resource Utilization Experiment (MOXIE) experiment has paved the way for

Mars exploration with little reliance on Earth [40]. Furthermore, reusable infrastructure elements and vehicles in

space can also drastically change future space exploration. For example, NASA’s Gateway concept is a representative

example of in-space infrastructure for supporting exploration of the Moon and beyond [41]. To achieve sustainable deep

space exploration, designing effective multi-mission space exploration campaigns leveraging such in-space logistics

infrastructure elements is critical.

Like in the ISAM application, there is a significant need for logistics-driven approaches to designing and planning

robotic and human space exploration campaigns. The key questions that this paper will focus on include:

• A2Q1: how to analyze the performance of a logistics strategy in the context of a space exploration campaign;

• A2Q2: how to plan and schedule the missions in a multi-mission exploration campaign with different types of

trajectories (high-thrust, low-thrust, etc.) leveraging in-space infrastructure;

• A2Q3: how to design and size the exploration vehicles and resource infrastructure technologies such as depots

and ISRU plants;

• A2Q4: how to respond to uncertainties in launch delay, infrastructure performance, etc. in a space campaign; and

• A2Q5: how to build the relationship between the governments and commercial players who supply in-space

infrastructure.

Logistics-driven research has played and will continue to play an important role in answering these questions.

3
C. Application 3 (A3): Mega-Scale Satellite Constellations

Another growing trend in low-Earth space is mega-scale satellite constellations. Many entities have proposed and

developed such systems including OneWeb [42], SpaceX’s Starlink [43], and Amazon’s Project Kuiper [44]. The

Department of Defense is also interested in concepts such as satellite swarms and proliferated low Earth orbit (pLEO).

Although satellite constellations have been developed and launched in the past, including Iridium [45] and Globalstar

[46], those conventional ones involved tens of satellites in each constellation. In contrast, those recent constellations

will involve hundreds to thousands of satellites, which is a significantly larger scale than any conventional constellation

system.

It is important to note that mega-scale constellations are not just a larger scale of small constellations. Instead, we

encounter new logistics challenges due to the scalability issue of conventional methods. The key questions include:

• A3Q1: how to launch and deploy a mega-scale constellation when it does not fit in a single launch vehicle;

• A3Q2: how to analyze the system performance and allocate the on-orbit spares for the constellation so that it

reliably satisfies performance requirements when failures would happen more frequently and in a more distributed

way than conventional satellite systems;

• A3Q3: how to flexibly reconfigure a satellite constellation when there is a change in the demand; and

• A3Q4: how to manage the commercial multi-stakeholder ecosystem for large-scale constellations.

As the scale and complexity of the space systems of interest grow, logistics-driven techniques will be critical to address

some of the inherent key challenges.

D. Mapping of Literature to Each Application Question

The following Table 1 shows the mapping of each study reviewed in this paper to each key research question listed

above. In our review, we primarily focus on the literature in recent years, but some earlier representative ones are

also included to provide the context. The technical details of each work are revisited as they are re-categorized by the

employed technical logistics-driven methods in Sec.III. Note that most of these questions are still open and new studies

are being performed as we speak.

4
 Table 1 Categorization of the State of the Art (SOTA) by Application Questions

Application Questions SOTA

Application 1: ISAM for Satellites
A1Q1: how to analyze the performance of an ISAM architecture [24–26, 47–49]
A1Q2: how to tactically plan and schedule the ISAM operations [50–56]
A1Q3: how to strategically architect the ISAM infrastructure elements such as depots and vehicles [54, 55, 57]
Application 2: Multi-Mission Space Exploration Campaigns
A2Q1: how to analyze the performance of a logistics strategy [58–68]
A2Q2: how to plan and schedule the missions in a multi-mission exploration campaign [35, 69–77]
A2Q3: how to design and size the exploration vehicles and resource infrastructure technologies [73, 78–81]
A2Q4: how to respond to uncertainties in launch delay, infrastructure performance, etc. [82, 83]
A2Q5: how to build the relationship between the governments and commercial players [84–87]
Application 3: Mega-Scale Satellite Constellations
A3Q1: how to launch and deploy a mega-scale constellation [88–90]
A3Q2: how to analyze the system performance and allocate the on-orbit spares for the constellation [91–94]
A3Q3: how to flexibly reconfigure a satellite constellation when there is a change in the demand [95, 96]
A3Q4: how to manage the commercial multi-stakeholder ecosystem for large-scale constellations [97–102]

III. Categorization of State of the Art by Logistics-Driven Methods

This section categorizes the same set of literature reviewed above by the logistics-driven methods used in each study

to aid the researchers in understanding the state of the art from the technical perspective.

Terrestrial logistics has a long history of research and practice, from which we have learned a lot to advance space

logistics research. This subsection summarizes three key method categories, each of which covers a subfield of logistics

research that is particularly relevant to space applications: (1) Network Flow Modeling and Optimization for Logistics

Planning and Scheduling; (2) Probabilistic Modeling and Queueing Theory for Logistics Performance Analysis; and (3)

Inventory Control for Resource Infrastructure Operations Management. Each subfield has been largely developed with

terrestrial applications in mind, and thus the developed theories and techniques cannot be directly applicable to space

applications. The challenges of applying terrestrial logistics-driven techniques are discussed for each subfield, as well

as how the literature has addressed some of these challenges. Lastly, the set of literature that extends beyond these

three key method categories in response to space-unique challenges is discussed. For more technical details on each

terrestrial logistics subfield, refer to the textbooks in the Operations Research (OR) field like Ref. [103, 104].

A. Method 1 (M1): Network Flow Modeling and Optimization for Logistics Planning and Scheduling

1. Motivation and Overview

As we consider how to manage a system of spacecraft and infrastructure distributed over the vast space in Earth’s

orbits, cislunar, and interplanetary space, it is natural to draw an analogy with how terrestrial logistics operations

manage their fleet of vehicles and commodity flows, for which network modeling and optimization has been a primary

5
tool. Figure 1 shows an example of network modeling for Earth-Moon-Mars exploration logistics.

Fig. 1 Example of network modeling for Earth-Moon-Mars exploration logistics [75].

A network-based approach models the system as a set of nodes and the arcs connecting them and employs various

techniques to analyze and optimize the commodity flow over that network. For example, for a space exploration

campaign problem, each orbital staging point (e.g., geosynchronous orbits, Lagrangian points, low-lunar orbits) is

modeled as a node and the trajectories connecting them are modeled as arcs. For an on-orbit servicing problem, the

nodes can also correspond to the customer satellites to be serviced. The commodities include everything that flows

over the network, including the propellant, spares, tools, payload, and vehicles themselves. The typical goal for such a

problem is to find a vehicle/commodity transportation routing plan and schedule to minimize a cost function, which can

be defined based on total campaign launch mass, cost, or other relevant metrics. Some of the well-known problem

types in network optimization include the traveling salesman problem (TSP), which finds the optimal route to visit

all nodes once and return to the origin node, and its generalized variant, the vehicle routing problem (VRP), which

finds the optimal set of routes for a fleet of vehicles to deliver to a given set of customers. Another relevant class of

network problem is the facility location problem (FLP), which finds the optimal location of the facilities such as depots

and servicing stations to optimally serve the distributed customers. Thus, we can draw analogies between a spacecraft

mission design problem and a TSP/VRP, and between the propellant depot or ISAM facility location problem with an

6
FLP.

2. Challenges in Space Logistics Applications

Unfortunately, optimizing and analyzing space logistics missions is substantially more complex than just solving a

conventional TSP/VRP/FLP over a ΔV map of orbits. There are some unique challenges that prevent a direct application

of existing terrestrial logistics approaches.

The most obvious difference between terrestrial logistics and space logistics is the orbital mechanics. Trajectory

beyond Earth’s orbits cannot be modeled as a two-body problem and thus, the generation of the ΔV or TOF will require a

large computational effort. For high-thrust trajectories, these quantities can be pre-computed (e.g., via Lambert problem

solvers) before optimizing the commodity flow, but this approach is not feasible for low-thrust trajectories, where the

ΔV and TOF can be flow-dependent; namely, if we assume a constant-thrust engine, the acceleration (and thus the ΔV

and TOF) depends on the mass of the commodities that the spacecraft carries. Thus, the logistics optimization problem

and the trajectory design problem are coupled. In addition, particularly when various trajectory options are considered

for logistics missions in cislunar space and beyond, drastically different time scales of the flight (from hours to months

or years) would need to be included in the same network, making the modeling of the time dimension challenging. On

top of that, the orbital mechanics also lead to complex time-dependent trajectory performance; for example, the relative

locations of the depots and servicers could be time-dependent, and thus cannot be pre-computed as a “map.” These

challenges are unique to space applications and make the conventional network optimization formulation ineffective or

infeasible.

Besides orbital mechanics, the nature of space infrastructure also poses additional unique challenges. In terrestrial

applications, when an infrastructure/facility is built and used, the deployment phase is typically relatively short compared

to the utilization phase of the infrastructure. However, that is not the case for space. As can be seen in examples like

the International Space Station (ISS), the deployment and assembly phase of infrastructure in space can be nearly

as long as its utilization phase; the boundary between deployment and operation is blurred. With ISRU or in-space

manufacturing, we can even use the resources generated by the previous stage of the infrastructure to deploy more

infrastructure (this concept is sometimes referred to as bootstrapping strategy [39]). Thus, we need to consider the

deployment of infrastructure together with its operations.

3. State of the Art

Fortunately, many (yet not all) of the above challenges have been tackled over 30 years of research in space logistics.

Network modeling and optimization are the most well-explored fields in space logistics modeling; see Fig. 1 as an

example. One of the first projects that leveraged network modeling for space mission design was SpaceNet [61], which

is a simulation and visualization software for space logistics operations. This project and its subsequent studies (e.g.,

7
Refs. [60, 62–66, 68]∗ ) primarily focused on supporting human space exploration campaigns with the then-active

Constellation Program in mind. Since then, network and graph-theoretic modeling have played a key role in analyzing

complex space logistics missions. Some of the representative works that focused on network modeling for space logistics

missions include Refs. [35, 70–72]. A more recent work specifically focused on the depot location problems for ISAM

applications with a facility location problem formulation [57]. A variant of network flow models, the assignment

problem, has also been used for satellite constellation deployment and reconfiguration analysis [90, 95, 96].

From the optimization perspective, a variety of different formulations have been proposed to efficiently and effectively

optimize the space exploration and/or ISAM mission design process. While some studies used metaheuristics such

as genetic algorithms or particle swarm optimization [51, 74] or simply evaluating all permutations of missions [53],

most studies developed network-based mixed-integer linear programming (MILP) formulations so that commercial

solvers such as Gurobi can solve them within guaranteed optimality gap. Note that, although the trajectory problem is a

nonlinear problem, a two-phase approach has been proposed to decouple the nonlinear trajectory problem from the

integer linear network problem [50, 52, 56, 90]. Two major types of network formulations for logistics optimization

include (1) a path formulation and (2) a node-arc formulation. A path formulation defines its decision variables as the

commodity flows over each possible vehicle’s and commodity’s path over the network, whereas a node-arc formulation

defines its decision variables as the vehicle/commodity flows over each arc. The choice of formulation depends on

the problem’s structure, constraints, and the solution method’s characteristics; for example, while a path formulation

would be effective in modeling a generic path-based cost and/or feasibility conditions, a node-arc formulation would

be more effective when the vehicle’s or commodity’s path is not easily definable or enumerable (e.g., reconfigurable

vehicles; ISRU resources). In the literature, path formulations have been typically developed in conjunction with

column-generation techniques; such examples include Martian surface vehicle routing [19] and lunar mission vehicle

routing (with no ISRU/ISAM) [69]. On the other hand, arc-based formulations have been developed when in-space

infrastructure elements are involved, such as multi-mission Mars-Moon-asteroid space campaign design with ISRU

[35, 71, 72]. Recently, another formulation, a path-arc formulation, was developed, which leverages the path formulation

for the vehicle and the node-arc formulation for the commodities [76].

To tackle the trajectory-related challenges, attempts have been made to develop approximate surrogate models

to compute the low-thrust, low-energy trajectory effectively leveraging Q-law and 3-body dynamics [77]. Such

approximation models can be incorporated into MILP optimization through piece-wise linear approximation so that

both (approximate) trajectory design and logistics optimization are performed concurrently.

To address the time-related challenges, the literature has also developed multiple techniques to model and optimize

the dynamic network flow. A classical method to consider the dynamic network flow is to use a time-expanded network,
∗ Some of the cited works (e.g., Ref. [62]) did not necessarily leverage the network structure explicitly, but they are included as they leveraged the

network modeling framework in SpaceNet.

8
where the nodes are copied over the time dimension with predefined time steps and network flow is optimized over

that expanded network. However, due to the significant time-scale differences in the problem and sometimes unknown

TOF without solving the logistics problem (i.e., in the low-thrust trajectories), a classical time-expanded network

can be ineffective or infeasible. To tackle this challenge, multiple variants of time-expanded networks have been

developed. For the case where the different time scales are known beforehand, a bi-scale time-expanded network was

developed [35]. For a specific type of problem such as Mars exploration leveraging the lunar resources, a partially-static

time-expanded network was also proposed as an approximation technique [35]. For a campaign that will largely repeat

once entering a steady state (e.g., resupply logistics for Mars habitats), a partially-periodic time-expanded network was

developed to effectively concurrently optimize both the build-up of the space system and its steady operations [75]. For

the flow-dependent TOFs, where any of the above variants of time-expanded networks would not be applicable, an

event-driven network was developed to optimize the logistics flow without specifying the time steps [77].

To consider the unique resource transformation mechanisms with ISRU and ISAM infrastructures, a new network

formulation was developed in the space logistics community: generalized multi-commodity network flow (GMCNF),

including its static [71] and dynamic versions [35]. This formulation generalizes the conventional multi-commodity

flow formulation so that it can model the commodity transformation along the arcs (including resource generation or

consumption); this general capability has enabled the modeling and optimization of missions with a significantly larger

variety of space infrastructure systems such as ISRU [35, 71] and ISAM systems [54, 55]. Combined with the partially

periodic time-expanded network mentioned above [75], we can also model the deployment of resource infrastructure

elements and their operations utilization concurrently.

B. Method 2 (M2): Probabilistic Modeling and Queueing Theory for Logistics Performance Analysis

1. Motivation and Overview

As we consider how to analyze the logistics performance of a space system under uncertainties, like a multi-satellite

system under geometrically distributed uncertain failures and a servicing infrastructure to respond to them, we can learn

from how fire stations or ambulance systems are designed to respond to uncertain demands distributed in a city. The

probabilistic modeling and spatial queueing theory have been very useful tools to that end. Figure 2 shows an example

of queueing modeling for an on-orbit servicing architecture.

Queueing theory is a mathematical theory to analyze the performance of a queue. With the service time distribution

and demand rate information as inputs, queueing theory analyzes the waiting time of a randomly arriving customer as

well as how long the queue is expected to be. The simplest queueing model, M/M/1, assumes a Poisson demand arrival

rate and an exponentially distributed service time and uses a Markov chain to model the system; more advanced models

have been developed such as a queue with multiple servers (e.g., M/M/m), a queue with capacity (e.g., M/M/1/c), a

queue with general service time distribution (e.g., M/G/1), and a queue with multiple classes of customers with different

9
 Fig. 2 Example of queueing modeling for an on-orbit servicing architecture [49].

priorities (priority queues). To model fire station or ambulance systems, spatial queueing models have been used, which

are based on the M/G/1 model with the service time distribution built based on the geometric relationship between the

servicer(s) and the potential customers [103]. Thus, analogously, to analyze the service responsiveness of an in-space

servicing system, for example, we can consider a similar spatial queueing model (i.e., orbital queueing model) with the

service time distribution derived based on the orbital relationship between the servicer(s) and the potential customers.

2. Challenges in Space Logistics Applications

Developing an orbital queueing model will involve challenges beyond a direct application of M/G/1 models. Each

of these challenges is not necessarily unexplored in terrestrial applications, but the combination of them, coupled with

orbital mechanics, makes the space application challenging.

One difference between the ISAM facility servicing and fire station or ambulance systems is the limited resources.

For any refueling and servicing operations, we need fuel and spares launched from the ground. Thus, we need to consider

an inventory management strategy for the storage space of the servicer and the depot in orbit if there is one. (The

study of inventory management is discussed in Sec. III.C.) This coupling between the queueing model and inventory

management makes the analysis challenging.

Additionally, the number of satellites a servicer serves is relatively limited. Thus, the demand distribution of the

satellite population would depend on the number of satellites that have failed and are being repaired (i.e., the less the

number of functioning satellites, the lower the failure rate for the entire system.); note that this feature can often be

ignored for terrestrial applications because of the large population that a fire station or an ambulance system serves.

A queue with a finite number of customers (often referred to as a finite-source queue) has been studied [105], but its

analysis is a substantially more challenging problem than conventional queues, particularly with generalized assumptions

about the service time, the number of servicers, the classes of customers with priorities, etc.

Furthermore, the dynamic nature of the demand and service rates can make the problem challenging. For example,

the demand rate (e.g., satellite failure rate) can be different depending on the phase of the satellite’s lifetime. In addition,

10
the required service time and cost are also time-dependent if the relative position of the servicing depots and the

customer satellites changes over time. An effective model needs to address these challenges so the system can be useful

for realistic space logistics analysis.

3. State of the Art

Modeling in-space logistics using probabilistic modeling techniques and queueing theory has been studied since the

1980s [47], but their theoretical developments for realistic space applications have been relatively limited.

Earlier studies focused on modeling ISAM applications based on queueing theory and analyzed them with discrete-

event simulations or agent-based simulations. Simulations have been used for on-orbit servicing analysis from the early

2000s [26]. More recently, attempts were made to model on-orbit servicing, recycling, and manufacturing applications

using discrete-event simulations [48, 67], in which the impacts of having different on-orbit recycling and manufacturing

technologies on the performance of on-orbit operations were quantitatively evaluated.

While simulations can tackle general scenarios, they are stochastic and computationally expensive, and thus not

effective for analyzing and optimizing the system; rather, there is a high need for analytic modeling of space logistics

systems. Various studies have employed probabilistic modeling techniques based on a Markov model for analyzing the

system availability of a constellation and designing its spare strategies [92, 93] or its servicing strategies [24, 25]. A

more recent study developed a semi-analytic model for an on-orbit servicing system based on the finite-source M/G/1

queueing model (i.e., M/G/1/K/K model) [49]. In this study, to model the limited launch resupply opportunities for the

spares from the ground, an inventory control technique was also integrated (see Sec.III.C.3). The resulting analytic

model matches well with the simulation model and can cut down the computational time from hours/days for simulations

down to seconds, enabling large-scale tradespace exploration and optimization in the early stage of the design.

C. Method 3 (M3): Inventory Control for Resource Infrastructure Operations Management

1. Motivation and Overview

As ISAM becomes a trend in space development, it is critical to explore the concepts of in-space propellant depots,

spare/tool warehouses, and other types of infrastructure. These infrastructure elements need to be regularly refueled or

resupplied from the ground (or possibly ISRU facilities on the Moon or Mars) and thus, the operations management of

the inventory can become a research question. To this end, we can draw an analogy between the inventory control of

in-space depots/warehouses and that of terrestrial depots/warehouses. Figure 3 shows an example of inventory modeling

for spare strategy analysis for a mega-scale satellite constellation.

Inventory management is a subfield in logistics that analyzes and optimizes inventory policies, including the timing

and quantity of orders, to minimize cost. Typically, the cost metric is the summation of the holding cost, purchase cost,

fixed ordering cost, and shortage penalty. Inventory management becomes different when there are uncertainties in the

11
Fig. 3 Example of inventory modeling for spare strategy analysis for a mega-scale satellite constellation [94].

demand and the lead time (i.e., the time between the order and its delivery). Given a cost metric, we can formulate an

optimization problem to find the optimal policy under uncertainties; one such problem is the Newsvendor model [104].

When there are constraints to the system on how the orders need to be made (e.g., periodic launch, launch capacity), we

can develop a parametric inventory policy and optimize their parameters. Some popular policies include (1) a periodic

review, order-up-to policy, which periodically orders to fill the inventory up to a certain level; (2) a reorder point, order

quantity policy, which orders a fixed quantity as soon as the inventory level drops below a threshold; and (3) a reorder

point, order-up-to policy, which orders to fill the inventory up to a certain level as soon as the inventory level drops below

a threshold. These parametric policies can be applied and optimized to space logistics context as well depending on the

structure of the problem. Furthermore, if we can consider multiple layers of inventory (e.g., manufacturer → warehouses

→ retailers), we can consider a multi-echelon inventory model; this can be useful to spacecraft spare inventory, where

Earth is modeled as manufacturer, spare orbits/depots as warehouses, and customer satellites as retailers.

2. Challenges in Space Logistics Applications

Applying inventory management to space logistics applications involves orbital mechanics, which often causes

additional challenges in the modeling of stochastic demand and lead time.

For the satellite refueling or repairing applications in the context of ISAM, the demand corresponds to the customer

satellites’ needs for fuels, materials, and spares. As reviewed in Sec. III.B, the demand depends on the phase of the

satellites’ lifetime and thus can be time-varying. Furthermore, if servicing operations are involved, the demand is

coupled with the transportation of the servicer (governed by the spatial queueing model). Thus, an integrated model

between queueing and inventory management is often needed.

In addition, a more unique challenge in space logistics application is the lead time distribution. The lead time

corresponds to the time between the order and its delivery, which is modeled based on orbital mechanics. However, due

to the dynamic nature of orbital mechanics, the relative position of the depots and customer satellites can vary over time.

12
Thus, when multiple depots and multiple customer satellites are considered, the satellites need to be resupplied from the

closest depot with a sufficient inventory level. This additional complexity in the modeling, which can be challenging if

complex orbital mechanics is involved, is one of the unique requirements for space applications.

3. State of the Art

The idea of inventory management has been conceived for satellite replenishment since the 1960s [91]. When space

logistics research was applied to human space exploration, inventory management was also recognized as an important

area [58, 59]. However, it was not until recently that mathematical inventory control theory was rigorously applied to

tackle realistic large-scale space logistics problems.

One of the recent studies that focused on developing an inventory control model is Ref. [94]. In this work, a

multi-echelon inventory model was developed for spare strategies for mega-scale constellations. In this work, separate

spare parking orbits are considered at a lower altitude than the original constellation orbit, and an analytic multi-echelon

inventory model based on a reorder point, order quantity policy was developed by modeling Earth as the manufacturer,

spare parking orbits as warehouses, and the constellation as the retailers. The resulting inventory model can be optimized

efficiently to identify the optimal number of parking orbits and their locations as well as their inventory control strategies.

Another work that optimized the inventory policy under uncertainties is Ref. [82]. In this work, a model was

developed to optimize space station logistics under launch schedule uncertainties. Specifically, the inventory policy (i.e.,

referred to as a decision rule) was parameterized by the safety stock level of the onboard inventory, and this parameter

was optimized along with the logistics decisions to balance the loss due to supply shortage and the cost of extra supply

(i.e., safety stock). Namely, it optimized the operational strategy to respond to the uncertainties in the launch delays.

In addition, as reviewed earlier, inventory control was also used in combination with queueing models in Ref. [49].

In this work, the inventory model is used to estimate how often a stockout happens (i.e., not enough spares are available

for further servicing missions) and how much delay is expected when it happens. This stochastic model is used as part

of the service time distribution model, which is then fed back into the queueing model to evaluate the performance of

the ISAM system.

D. Extensions of Basic Models

Beyond the above basic models introduced, the literature has explored the extensions of these basic methods to

address unique challenges in space logistics challenges. Some are reviewed below.

1. Concurrent Design of Vehicle/Infrastructure and Space Logistics Network

One unique feature of space missions is that we often design space vehicles and infrastructure elements dedicated

to a certain space campaign. Thus, we also have an interesting research question in space logistics regarding how to

13
achieve the concurrent design optimization of the vehicle, infrastructure, and logistics commodity flow network. This

problem often involves nonlinearity and multi-disciplinary design analysis due to the nature of the spacecraft and space

infrastructure design. Thus, to solve such a problem, we need new strategic formulations and solution methods beyond

traditional logistics research.

To enable the incorporation of vehicle sizing and infrastructure design, a few techniques have been developed

depending on the required fidelity. If a linear vehicle and infrastructure sizing model is sufficient, a regular MILP

formulation can be used [35, 71, 72]. A higher-fidelity version would be to use a more piece-wise linear approximation

for the vehicle sizing, which can still be converted into a MILP formulation but requires more variables and constraints

[73]. For a yet higher-fidelity case where a nonlinear model needs to be used, an embedded optimization [78] was

developed earlier, and a more advanced and effective Augmented Lagrangian Coordination (ALC) approach has been

developed recently [81]. The ALC-based approach is based on multi-disciplinary optimization and leverages the

unique structure of the problem by decomposing the whole mixed-integer nonlinear programming problem into a set of

mixed-integer quadratic programming problems and nonlinear programming problems, each of which can be solved

using specialized solvers. Finally, to incorporate the multi-subsystem interaction within the ISRU or other in-space

infrastructure into the logistics optimization, a multi-fidelity network flow was developed to capture these interactions

effectively [79, 80].

2. Incorporating Operational Uncertainties in Space Logistics Network Optimization

Although earlier space logistics research has primarily focused on deterministic problem setting, recent studies

started to incorporate uncertainties such as launch delays, demand fluctuation, and infrastructure performance decay.

A major approach is to characterize the uncertainties into a set of possible scenarios and optimize the strategy

to respond to these scenarios together with the baseline design. The most straightforward formulation to this end is

stochastic optimization, which has been applied to multi-stage mega-scale constellation deployment under demand

uncertainties [88, 89]. This approach is simple but cannot effectively provide a strategy to respond to the uncertainties.

As an alternative, as reviewed earlier, the decision-rule-based optimization was developed where the strategies (i.e.,

decision rules) to respond to the uncertainties are parametrically defined and optimized (e.g., safety stock to prepare for

launch delays) [82]. Furthermore, to handle more generic uncertainties in a space campaign, where the decision rules

cannot be intuitively parameterized, a Hierarchical Reinforcement Learning (HRL) framework was developed [83].

Leveraging the strength of reinforcement learning (RL) to handle uncertainties and MILP’s strength to handle a large

variable space well, the HRL framework uses RL to provide high-level guidance and uses MILP to optimize the detailed

action to achieve that guidance. This success was one of the first studies that incorporated machine learning into space

logistics.

Another approach to determining how to respond to uncertainties is to reoptimize as soon as we learn more

14
information about the uncertainties [51, 53]. Building upon this idea, a rolling horizon approach was developed for

ISAM applications to respond to newly obtained demand information over time [54, 55] and a reconfiguration approach

was developed for satellite constellation configuration modification as we acquire new demand [95, 96].

3. Space Logistics Modeling for Commercialization

Most existing works have focused on the centralized optimization of the entire space logistics network. However,

given the recent trend of space commercialization, this is not necessarily the best modeling approach. Thus, an interesting

recent extension of the above work is to incorporate multiple players in space logistics. To this end, Ref. [87] extended

network optimization research with the game theory to optimize the incentive that the government should provide to the

commercial players (e.g., infrastructure providers). Other works used alternative approaches to examine the affordability

and commercialization of space missions [85] and investigated a method to use the commercial suitability as an explicit

criterion of the initial architecture selection [86]. Furthermore, various studies developed an interactive game as well as

a simulation framework to assess the multi-stakeholder space systems [84, 97–102].

IV. Summary of Reviewed Literature

Table 2 summarizes how each reviewed literature is mapped to the corresponding application area and logistics-driven

method. Note that some papers span over multiple methods (e.g., Ref. [49]). While not all combinations of the

application areas and logistics-driven methods have an equal amount of room for exploration, this table shows some

well-explored and under-explored combinations of application areas and methods.

Table 2 Mapping of the Applications and Methods

Applications/Methods M1 M2 M3 Extensions
A1: ISAM for Satellites [50–57] [24–26, 47–49] [49] [51, 53–55]
A2: Space Exp. Campaigns [35, 60–66, 68–72, 74–77] [67] [58, 59, 82] [35, 71–73, 78–87]
A3: Satellite Constellations [90, 95, 96] [92, 93] [91, 94] [88, 89, 95–102]

In addition, Figure 4 shows the number of citations (as of Feb. 20, 2024; according to Google Scholar) for each

reviewed paper. The data is split into three time frames, the 2000s or earlier, the 2010s, and the 2020s, to show the most

impactful papers in each decade in the space logistics area. Each reference also has its application (A1/A2/A3) and

method(s) (M1/M2/M3/Ext) listed, where Ext stands for one of the extensions. Although the number of citations does

not necessarily indicate the value of the work, this figure shows the trend and the shift of focus over time in terms of

both the applications and methods.

While many questions have been addressed by space logistics research, there is enormous room for growth. Some

examples of the open fundamental research questions are: (1) scalable algorithms (including heuristics) for large-scale

space logistics network optimization: (2) complex queueing networks for space logistics analysis; and (3) dynamic

15
 2009 or Before
300
A1: ISAM for Satellites
Number of citations

250
A2: Space Exp. Campaigns
200 A3: Satellite Constellations
150
100
50
0
[88] [26] [24] [95] [25] [69] [61] [60] [58] [62] [92] [93] [59] [91] [78] [47]
Ext M2 M2 M1/ M2 M1 M1 M1 M3 M1 M2 M2 M3 M3 Ext M2
Ext
2010-2020
120
Number of citations

100
80
60
40
20
0
[71] [35] [73] [66] [72] [89] [99] [70] [52] [94] [75] [98] [50] [51] [48] [97] [64] [63] [74] [65] [84] [85] [76]
M1/ M1/ Ext M1 M1/ Ext Ext M1 M1 M3 M1 Ext M1 M1/ M2 Ext M1 M1 M1 M1 Ext Ext M1
Ext Ext Ext Ext
2020-2023
40
Number of citations

30

20

10

0
[79] [53] [49] [67] [54] [80] [77] [83] [82] [81] [55] [96] [87] [90] [86] [57] [56] [68] [101][100][102]
Ext M1/ M2/ M2 M1/ Ext M1 Ext M3/ Ext M1/ M1/ Ext M1 Ext M1 M1 M1 Ext Ext Ext
Ext M3 Ext Ext Ext Ext
References and Used Methods

Fig. 4 Number of citations (as of Feb. 20, 2024; according to Google Scholar) of the reviewed papers in each
decade.

multi-item inventory for space logistics applications, among others. Furthermore, given the nature of the growing

space logistics applications, we can expect to see logistics-driven methods applied to more problems beyond the ones

reviewed above. Drawing analogies between more terrestrial logistics applications and space logistics applications

can lead to new research questions. Finally, most space logistics works have assumed simplified trajectory, vehicle,

and operational models to enable large-scale tradespace exploration and optimization; once the solutions are obtained

from these low-/mid-fidelity models, we need higher-fidelity simulations to validate these results. Such validations are

16
critical yet largely unexplored research areas.

V. Conclusion
This paper provides an overview of the state of the art for space logistics modeling and optimization. The recent

literature is categorized in two ways: (1) by application questions that are addressed; and (2) by logistics-driven methods

that are used in the studies. The applications reviewed in this paper are determined by the needs of the community; the

specific applications of interest include: (1) ISAM for satellites; (2) multi-mission space exploration campaigns; and (3)

mega-scale satellite constellations. The logistics-driven methods reviewed in this paper are determined by mapping

the relevant major logistics research subfield to space applications; the connection between each logistics research

subfield to space logistics applications is reviewed and the relevant unique challenges in space applications are discussed.

The reviewed logistics-driven methods include: (1) network flow modeling and optimization for logistics planning

and scheduling; (2) probabilistic modeling and queueing theory for logistics performance analysis; and (3) inventory

control for resource infrastructure operations management. Several extensions of these basic models that tackled

unique challenges in space applications are also reviewed. We expect that the first application-based categorization

helps practitioners understand the existing research that can answer their application questions, whereas the second

method-based categorization helps the researchers to understand the technical perspective of the literature and identify

new under-explored research directions.

Acknowledgment
The author thanks the members of the AIAA Space Logistics Technical Committee for their helpful comments. The

author also thanks Masafumi Isaji and Yuri Shimane for their suggestions about data visualization and improvement of

the paper. While the initial version of the work was not associated with a funding source, its refinement and revision

was conducted with support from the Air Force Office of Scientific Research (AFOSR), as part of the Space University

Research Initiative (SURI), under award number FA9550-23-1-0723.

Author Biography
Dr. Koki Ho is the Dutton-Ducoffe Professor, Associate Professor, and the director of the Space Systems Optimization

Group in the Daniel Guggenheim School of Aerospace Engineering at Georgia Institute of Technology. His research

focuses on developing modeling and optimization methods for rigorous space mission analysis and design. His unique

research connecting logistics-based modeling, optimization, systems engineering, and space applications has provided a

substantial impact on modern and future space missions that involve multiple missions, multiple vehicles, and reusable

infrastructure elements. Dr. Ho earned his Ph.D. at the Massachusetts Institute of Technology and his bachelor’s and

master’s degrees at the University of Tokyo. He is the recipient of the NSF CAREER Award (2020), the NASA Early

17
Career Faculty Award (2019), the DARPA Young Faculty Award (2019), and the Luigi Napolitano Award (2015),

and he is a co-author of one of the most downloaded Acta Astronautica articles. Dr. Ho is the Chair of the AIAA

Space Logistics Technical Committee (2017–2024) and also serves on the Steering Committee for the NASA-funded

Consortium for Space Mobility and ISAM Capabilities (COSMIC).

References
[1] Raymond, J. W., “Spacepower: Doctrine for Space Forces,” Technical Report, Special Report, 2020.

[2] AIAA Space Logistics Technical Committee, “AIAA Space Logistics Technical Committee,” Web page, [Accessed on May 11,

2023]. URL https://www.aiaa-sltc.org/.

[3] Freeman, R. J., Gogerty, D. C., Graves, G. W., and Brooks, R. B., “A mathematical model of supply support for space

operations,” Operations Research, Vol. 14, No. 1, 1966, pp. 1–15. https://doi.org/10.1287/opre.14.1.1.

[4] Root, M., “An evaluation of the national space inventory for logistics,” AIAA 3rd Annual Meeting, 1966, p. 866.

[5] Carrillo, M. J., Jacobsen, S. E., Abell, J. B., and Lippiatt, T. F., A development of logistics management models for the Space

Transportation System, Vol. 3083, 1983.

[6] AIAA, “AIAA/SOLE 1st Space Logistics Symposium,” , [Accessed on Nov 21, 2023]. URL https://arc.aiaa.org/doi/book/10.

2514/MSLS87.

[7] Shepard, K., “The role of inventory management in satellite servicing,” AIAA/SOLE 1st Space Logistics Symposium, 1987, p.

667. https://doi.org/10.2514/6.1987-667.

[8] Bell, J., and McGeehan, R., “Logistics and operations integration requirements to support Space Station servicing of free

flying spacecraft-OMV flight operation,” AIAA/SOLE 1st Space Logistics Symposium, 1987, p. 686. https://doi.org/https:

//doi.org/10.2514/6.1987-686.

[9] Barry, J., “Application of artificial intelligence (AI) concepts for employing logistics resources in the SDI environment,”

AIAA/SOLE 1st Space Logistics Symposium, 1987, p. 692. https://doi.org/10.2514/6.1987-692.

18
[10] Poliskie, M., The Planning and Execution of Human Missions to the Moon and Mars, American Institute of Aeronautics and

Astronautics, Inc., 2023. https://doi.org/10.2514/4.106545.

[11] Shishko, R., and de Weck, O. L., “Interplanetary Supply Chains and Space Logistics: Enabler of the Final Frontier,” The

Planning and Execution of Human Missions to the Moon and Mars, American Institute of Aeronautics and Astronautics, Inc.

Reston, Virginia, 2023, pp. 151–194. https://doi.org/10.2514/5.9781624106545.0151.0194.

[12] Ho, K., Chen, H., and Du Jonchay, T. S., “Mathematical Methods for Space Mission Planning and Architecture Design,” The

Planning and Execution of Human Missions to the Moon and Mars, American Institute of Aeronautics and Astronautics, Inc.

Reston, Virginia, 2023, pp. 597–624. https://doi.org/10.2514/5.9781624106545.0597.0624.

[13] Siddiqi, A., and de Weck, O. L., “Spare parts requirements for space missions with reconfigurability and commonality,”

Journal of Spacecraft and Rockets, Vol. 44, No. 1, 2007, pp. 147–155. https://doi.org/10.2514/1.21847.

[14] Owens, A., de Weck, O. L., Stromgren, C., Goodliff, K. E., and Cirillo, W., “Supportability challenges, metrics, and

key decisions for future human spaceflight,” AIAA SPACE and Astronautics Forum and Exposition, 2017, p. 5124.

https://doi.org/10.2514/6.2017-5124.

[15] Kline, R. C., and Bachman, T. C., “Estimating spare parts requirements with commonality and redundancy,” Journal of

Spacecraft and Rockets, Vol. 44, No. 4, 2007, pp. 977–984. https://doi.org/10.2514/1.28072.

[16] Siddiqi, A., de Weck, O., and Iagnemma, K., “Reconfigurability in planetary surface vehicles: modeling approaches and case

study,” Journal of the British Interplanetary Society, Vol. 59, No. 12, 2006, pp. 450–460.

[17] Ahn, J., de Weck, O. L., and Hoffman, J., “An Optimization Framework for Global Planetary Surface Exploration Campaigns,”

Journal of the British Interplanetary Society, Vol. 61, No. 12, 2008, pp. 487–498.

[18] Ahn, J., and de Weck, O. L., “Mars surface exploration caching from an orbiting logistics depot,” Journal of Spacecraft and

Rockets, Vol. 47, No. 4, 2010, pp. 690–694. https://doi.org/10.2514/1.37694.

[19] Ahn, J., de Weck, O. L., Geng, Y., and Klabjan, D., “Column generation based heuristics for a generalized location routing

problem with profits arising in space exploration,” European Journal of Operational Research, Vol. 223, 2012, pp. 47–59.

https://doi.org/10.1016/j.ejor.2012.06.018.

[20] Lee, D., and Ahn, J., “Integrated optimization of planetary rover layout and exploration routes,” Engineering Optimization,

Vol. 50, No. 1, 2018, pp. 164–182. https://doi.org/10.1080/0305215X.2017.1299718.

[21] Lee, D., and Ahn, J., “Vehicle routing problem with vector profits with max-min criterion,” Engineering Optimization, Vol. 51,

No. 2, 2019, pp. 352–367. https://doi.org/10.1080/0305215X.2018.1458845.

[22] Choi, E., and Ahn, J., “Optimal planetary surface routing considering exploration values and synergies,” Journal of Aerospace

Information Systems, Vol. 19, No. 6, 2022, pp. 406–420. https://doi.org/10.2514/1.I011052.

19
[23] Choi, E., and Ahn, J., “Optimal planetary surface exploration with heterogeneous agents,” Journal of Spacecraft and Rockets,

Vol. 60, No. 4, 2023, pp. 1343–1354. https://doi.org/10.2514/1.A35632.

[24] Saleh, J. H., Lamassoure, E., and Hastings, D. E., “Space systems flexibility provided by on-orbit servicing: Part 1,” Journal

of Spacecraft and Rockets, Vol. 39, No. 4, 2002, pp. 551–560. https://doi.org/10.2514/2.3844.

[25] Lamassoure, E., Saleh, J. H., and Hastings, D. E., “Space systems flexibility provided by on-orbit servicing: Part 2,” Journal

of Spacecraft and Rockets, Vol. 39, No. 4, 2002, pp. 561–570. https://doi.org/10.2514/2.3845.

[26] Long, A. M., Richards, M. G., and Hastings, D. E., “On-orbit servicing: a new value proposition for satellite design and

operation,” Journal of Spacecraft and Rockets, Vol. 44, No. 4, 2007, pp. 964–976. https://doi.org/10.2514/1.27117.

[27] Galabova, K. K., and de Weck, O. L., “Economic case for the retirement of geosynchronous communication satellites via

space tugs,” Acta Astronautica, Vol. 58, No. 9, 2006, pp. 485–498. https://doi.org/10.1016/j.actaastro.2005.12.014.

[28] Gralla, W. L., and de Weck, O. L., “Strategies for on-orbit assembly of modular spacecraft,” Journal of the British Interplanetary

Society, Vol. 60, No. 6, 2007, pp. 219–227.

[29] DARPA, “DARPA Robotic Servicing of Geosynchronous Satellites (RSGS) Program,” [Accessed on May 11, 2023]. URL

https://www.darpa.mil/program/robotic-servicing-of-geosynchronous-satellites.

[30] Northrop Grumann, “Northrop Grumman SpaceLogistics,” [Accessed on May 11, 2023]. URL https://www.northropgrumman.

com/space/space-logistics-services/.

[31] DARPA, “DARPA Novel Orbital and Moon Manufacturing, Materials, and Mass-efficient Design (NOM4D) Program,”

[Accessed on May 11, 2023]. URL https://www.darpa.mil/program/novel-orbital-and-moon-manufacturing-materials-and-

mass-efficient-design.

[32] Gaebler, J. A., Lugo, R. A., Axdahl, E. L., Chai, P. R., Grimes, M. S., Long, M. R., Rowland, R. A., and Wilhite, A., “Reusable

lunar transportation architecture utilizing orbital propellant depots,” AIAA Space 2009 Conference and Exposition, American

Institute of Aeronautics and Astronautics Inc., 2009. https://doi.org/10.2514/6.2009-6711.

[33] Wilhite, A., Arney, D. C., and Chai, P., “Permanent manned outpost with commercial launch and propellant depots,” AIAA

SPACE Conference and Exposition 2012, 2012. https://doi.org/10.2514/6.2012-5130.

[34] Noevere, A. T., and Wilhite, A., “Techniques for integrated thermal-structural design of an orbital propellant depot,”

AIAA SPACE 2013 Conference and Exposition, American Institute of Aeronautics and Astronautics Inc., 2013. https:

//doi.org/10.2514/6.2013-5495.

[35] Ho, K., de Weck, O. L., Hoffman, J. A., and Shishko, R., “Dynamic modeling and optimization for space logistics using

time-expanded networks,” Acta Astronautica, Vol. 105, 2014, pp. 428–443. https://doi.org/10.1016/j.actaastro.2014.10.026.

[36] NASA, “NASA ISRU,” Web page, [Accessed on May 11, 2023]. URL https://www.nasa.gov/isru.

20
[37] Shishko, R., Fradet, R., Saydam, S., Tapia-Cortez, C., Dempster, A., Coulton, J., and Do, S., “An integrated economics model

for ISRU in support of a Mars colony–Initial results report,” 10th Symposium on Space Resource Utilization, 2017, p. 0422.

https://doi.org/10.2514/6.2017-0422.

[38] Shishko, R., Fradet, R., Do, S., Saydam, S., Tapia-Cortez, C., Dempster, A. G., and Coulton, J., “Mars Colony in situ

resource utilization: An integrated architecture and economics model,” Acta Astronautica, Vol. 138, 2017, pp. 53–67.

https://doi.org/10.1016/j.actaastro.2017.05.024.

[39] Kornuta, D., Abbud-Madrid, A., Atkinson, J., Barr, J., Barnhard, G., Bienhoff, D., Blair, B., Clark, V., Cyrus, J., DeWitt,

B., Dreyer, C., Finger, B., Goff, J., Ho, K., Kelsey, L., Keravala, J., Kutter, B., Metzger, P., Montgomery, L., Morrison,

P., Neal, C., Otto, E., Roesler, G., Schier, J., Seifert, B., Sowers, G., Spudis, P., Sundahl, M., Zacny, K., and Zhu, G.,

“Commercial lunar propellant architecture: A collaborative study of lunar propellant production,” REACH, Vol. 13, 2019.

https://doi.org/10.1016/j.reach.2019.100026.

[40] Hecht, M., Hoffman, J., Rapp, D., McClean, J., SooHoo, J., Schaefer, R., Aboobaker, A., Mellstrom, J., Hartvigsen,

J., Meyen, F., et al., “Mars oxygen ISRU experiment (MOXIE),” Space Science Reviews, Vol. 217, 2021, pp. 1–76.

https://doi.org/https://doi.org/10.1007/s11214-020-00782-8.

[41] NASA, “NASA Gateway Program,” Web page, [Accessed on May 11, 2023]. URL https://www.nasa.gov/gateway.

[42] OneWeb, “OneWeb,” Web page, [Accessed on May 11, 2023]. URL https://oneweb.net/.

[43] SpaceX, “SpaceX Starlink Constellation,” Web page, [Accessed on May 11, 2023]. URL https://www.starlink.com/.

[44] Amazon, “Amazon Project Kuiper,” [Accessed on May 11, 2023]. URL https://www.aboutamazon.com/what-we-do/devices-

services/project-kuiper.

[45] Fossa, C., Raines, R., Gunsch, G., and Temple, M., “An overview of the IRIDIUM (R) low Earth orbit (LEO) satellite system,”

Proceedings of the IEEE 1998 National Aerospace and Electronics Conference. NAECON 1998. Celebrating 50 Years (Cat.

No.98CH36185), 1998, pp. 152–159. https://doi.org/10.1109/NAECON.1998.710110.

[46] Dietrich, F. J., Metzen, P., and Monte, P., “The Globalstar cellular satellite system,” IEEE Transactions on Antennas and

propagation, Vol. 46, No. 6, 1998, pp. 935–942. https://doi.org/10.1109/8.686783.

[47] Martin, D., II, “Application of queuing theory to on-orbit logistics,” 2nd Space Logistics Symposium, 1988, p. 4752.

https://doi.org/10.2514/6.1988-4752.

[48] Sears, P. J., and Ho, K., “Impact evaluation of in-space additive manufacturing and recycling technologies for on-orbit

servicing,” Journal of Spacecraft and Rockets, Vol. 55, 2018, pp. 1498–1508. https://doi.org/10.2514/1.A34135.

[49] Ho, K., Wang, H., Trempe, P. A. D., Tristan Sarton du Jonchay, T., and Tomita, K., “Semi-analytical model for design

and analysis of on-orbit servicing architecture,” Journal of Spacecraft and Rockets, Vol. 57, 2020, pp. 1129–1138.

https://doi.org/10.2514/1.A34663.

21
[50] Bang, J., and Ahn, J., “Two-phase framework for near-optimal multi-target Lambert rendezvous,” Advances in Space Research,

Vol. 61, 2018, pp. 1273–1285. https://doi.org/10.1016/j.asr.2017.12.025.

[51] Verstraete, A. W., Anderson, D., Louis, N. M. S., and Hudson, J., “Geosynchronous earth orbit robotic servicer mission

design,” Journal of Spacecraft and Rockets, Vol. 55, 2018, pp. 1444–1452. https://doi.org/10.2514/1.A33945.

[52] Bang, J., and Ahn, J., “Multitarget rendezvous for active debris removal using multiple spacecraft,” Journal of Spacecraft and

Rockets, Vol. 56, 2019, pp. 1237–1247. https://doi.org/10.2514/1.A34344.

[53] Hudson, J. S., and Kolosa, D., “Versatile on-orbit servicing mission design in geosynchronous earth orbit,” Journal of

Spacecraft and Rockets, Vol. 57, 2020, pp. 844–850. https://doi.org/10.2514/1.A34701.

[54] Sarton du Jonchay, T., Chen, H., Gunasekara, O., and Ho, K., “Framework for modeling and optimization of on-orbit

servicing operations under demand uncertainties,” Journal of Spacecraft and Rockets, Vol. 58, 2021, pp. 1157–1173.

https://doi.org/10.2514/1.A34978.

[55] Sarton du Jonchay, T., Chen, H., Isaji, M., Shimane, Y., and Ho, K., “On-orbit servicing optimization framework with high- and

low-thrust propulsion tradeoff,” Journal of Spacecraft and Rockets, Vol. 59, 2022, pp. 33–48. https://doi.org/10.2514/1.A35094.

[56] Lee, D., and Ahn, J., “Optimal Multitarget Rendezvous Using Hybrid Propulsion System,” Journal of Spacecraft and Rockets,

Vol. 60, No. 2, 2023, pp. 689–698. https://doi.org/10.2514/1.A35540.

[57] Shimane, Y., Gollins, N., and Ho, K., “Orbital facility location problem for satellite constellation servicing depots,” Journal of

Spacecraft and Rockets, Vol. 61, No. 3, 2024, pp. 808–825. https://doi.org/https://doi.org/10.2514/1.A35691.

[58] Shull, S., Gralla, E., Siddiqi, A., de Weck, O., and Shishko, R., “The future of asset management for human space exploration:

Supply classification and an integrated database,” Space 2006, 2006, p. 7232. https://doi.org/10.2514/6.2006-7232.

[59] Shull, S. A., Gralla, E. L., Silver, M., and de Weck, O. L., “Logistics information systems for human space exploration: State

of the art and emerging technologies,” SpaceOps 2006, 2006. https://doi.org/10.2514/6.2006-5733.

[60] Gralla, E., Shull, S., and de Weck, O. L., “A modeling framework for interplanetary supply chains,” Space 2006, 2006, p.

7229. https://doi.org/10.2514/6.2006-7229.

[61] de Weck, O. L., Simchi-Levi, D., Shishko, R., Ahn, J., Gralla, E., Klabjan, D., Mellein, J., Shull, A., Siddiqi, A., Bairstow, B.,

et al., “Spacenet v1. 3 user’s guide,” NASA/TP, Vol. 214725, 2007, p. 2007.

[62] Siddiqi, A., de Weck, O. L., Lee, G., and Shull, S., “Matrix modeling methods for spaceflight campaign logistics analysis,”

Journal of Spacecraft and Rockets, Vol. 46, No. 5, 2009, pp. 1037–1048. https://doi.org/10.2514/1.43319.

[63] Ishimatsu, T., Grogan, P., and de Weck, O. L., “Interplanetary trajectory analysis and logistical considerations of human Mars

exploration,” Journal of Cosmology, Vol. 12, 2010, pp. 3588–3600.

22
[64] Grogan, P., Siddiqi, A., and de Weck, O. L., “Matrix methods for optimal manifesting of multi-node space exploration systems,”

Journal of Spacecraft and Rockets, Vol. 48, No. 4, 2011, pp. 679–690. https://doi.org/10.2514/1.51870.

[65] Grogan, P., de Weck, O. L., and Lee, C., “Comparative usability study of two space logistics analysis tools,” AIAA SPACE

2011 Conference & Exposition, 2011, p. 7345. https://doi.org/10.2514/6.2011-7345.

[66] Grogan, P., Yue, H., and de Weck, O. L., “Space logistics modeling and simulation analysis using SpaceNet: Four application

cases,” AIAA Space 2011 Conference & Exposition, 2011, p. 7346. https://doi.org/10.2514/6.2011-7346.

[67] Moraguez, M., and de Weck, O., “Benefits of in-space manufacturing technology development for human spaceflight,” 2020

IEEE Aerospace Conference, IEEE, 2020, pp. 1–11. https://doi.org/10.1109/AERO47225.2020.9172304.

[68] Capra, L., Hilton, J., Bentley, S., Sherman, T., Alfaro, A., Savin, R., de Weck, O. L., and Grogan, P. T., “SpaceNet

Cloud: Web-based Modeling and Simulation Analysis for Space Exploration Logistics,” ASCEND 2021, 2021, p. 4068.

https://doi.org/10.2514/6.2021-4068.

[69] Taylor, C., Song, M., Klabjan, D., ; de Weck, O. L., and Simchi-Levi, D., “A mathematical model for interplanetary logistics,”

Logistics Spectrum, Vol. 41, 2007.

[70] Arney, D. C., and Wilhite, A. W., “Modeling space system architectures with graph theory,” Journal of Spacecraft and Rockets,

Vol. 51, 2014, pp. 1413–1429. https://doi.org/10.2514/1.A32578.

[71] Ishimatsu, T., de Weck, O. L., Hoffman, J. A., and Ohkami, Y., “Generalized multicommodity network flow model for the earth-

moon-mars logistics system,” Journal of Spacecraft and Rockets, Vol. 53, 2016, pp. 25–38. https://doi.org/10.2514/1.A33235.

[72] Ho, K., de Weck, O. L., Hoffman, J. A., and Shishko, R., “Campaign-level dynamic network modelling for spaceflight logistics

for the flexible path concept,” Acta Astronautica, Vol. 123, 2016, pp. 51–61. https://doi.org/10.1016/j.actaastro.2016.03.006.

[73] Chen, H., and Ho, K., “Integrated space logistics mission planning and spacecraft design with mixed-integer nonlinear

programming,” Journal of Spacecraft and Rockets, Vol. 55, 2018, pp. 365–381. https://doi.org/10.2514/1.A33905.

[74] Jagannatha, B. B., and Ho, K., “Optimization of in-space supply chain design using high-thrust and low-thrust propulsion

technologies,” Journal of Spacecraft and Rockets, Vol. 55, 2018, pp. 648–659. https://doi.org/10.2514/1.A34042.

[75] Chen, H., Lee, H. W., and Ho, K., “Space transportation system and mission planning for regular interplanetary missions,”

Journal of Spacecraft and Rockets, Vol. 56, 2019, pp. 12–20. https://doi.org/10.2514/1.A34168.

[76] Mcbrayer, K., Daniel, N., Deguignet, M., Velte, A., Robertson, B., and Mavris, D. N., “A network flow-based formulation to

optimize campaign alternatives for a reference lunar surface base,” International Astronautical Congress, 2019, pp. 21–25.

[77] Jagannatha, B. B., and Ho, K., “Event-driven network model for space mission optimization with high-thrust and low-thrust

spacecraft,” Journal of Spacecraft and Rockets, Vol. 57, 2020, pp. 446–463. https://doi.org/10.2514/1.A34628.

23
[78] Taylor, M. C., de Weck, P. O., and Klabjan, P. D., “Integrated Transportation System Design for Space Exploration Logistics,”

57th International Astronautical Congress, 2006, pp. D1–3.

[79] Chen, H., Sarton du Jonchay, T., Hou, L., and Ho, K., “Integrated in-situ resource utilization system design and logistics for

Mars exploration,” Acta Astronautica, Vol. 170, 2020, pp. 80–92. https://doi.org/10.1016/j.actaastro.2020.01.031.

[80] Chen, H., Sarton du Jonchay, T., Hou, L., and Ho, K., “Multifidelity space mission planning and infrastructure design framework

for space resource logistics,” Journal of Spacecraft and Rockets, Vol. 58, 2021, pp. 538–551. https://doi.org/10.2514/1.A34666.

[81] Isaji, M., Takubo, Y., and Ho, K., “Multidisciplinary design optimization approach to integrated space mission planning and

spacecraft design,” Journal of Spacecraft and Rockets, Vol. 59, 2022, pp. 1660–1670. https://doi.org/10.2514/1.A35284.

[82] Chen, H., Gardner, B. M., Grogan, P. T., and Ho, K., “Flexibility management for space logistics via decision rules,” Journal

of Spacecraft and Rockets, Vol. 58, 2021, pp. 1314–1324. https://doi.org/10.2514/1.A34985.

[83] Takubo, Y., Chen, H., and Ho, K., “Hierarchical reinforcement learning framework for stochastic spaceflight campaign design,”

Journal of Spacecraft and Rockets, Vol. 59, 2022, pp. 421–433. https://doi.org/10.2514/1.A35122.

[84] Grogan, P., and de Weck, O., “Federated simulation and gaming framework for a decentralized space-based resource economy,”

Earth and Space 2012: Engineering, Science, Construction, and Operations in Challenging Environments, 2012, pp.

1468–1477.

[85] Shishko, R., Price, H., Wilcox, B., Stoica, A., Howe, S., and Elliott, J., “An affordable lunar architecture emphasizing

commercial and international partnering opportunities,” 2018 IEEE Aerospace Conference, IEEE, 2018, pp. 1–16.

https://doi.org/10.1109/AERO.2018.8396423.

[86] Sarton du Jonchay, T., Chen, H., Wieger, A., Szajnfarber, Z., and Ho, K., “Space architecture design for commercial

suitability: A case study in in-situ resource utilization systems,” Acta Astronautica, Vol. 175, 2020, pp. 45–50. https:

//doi.org/10.1016/j.actaastro.2020.05.012.

[87] Chen, H., Ornik, M., and Ho, K., “Space exploration architecture and design framework for commercialization,” Journal of

Spacecraft and Rockets, Vol. 59, 2022, pp. 538–551. https://doi.org/10.2514/1.A35077.

[88] de Weck, O. L., de Neufville, R., and Chaize, M., “Staged deployment of communications satellite constellations in

low earth orbit,” Journal of Aerospace Computing, Information, and Communication, Vol. 1, No. 3, 2004, pp. 119–136.

https://doi.org/10.2514/1.6346.

[89] Lee, H. W., Jakob, P. C., Ho, K., Shimizu, S., and Yoshikawa, S., “Optimization of satellite constellation deployment strategy

considering uncertain areas of interest,” Acta Astronautica, Vol. 153, 2018, pp. 213–228. https://doi.org/10.1016/j.actaastro.

2018.03.054.

[90] Sung, T., and Ahn, J., “Optimal deployment of satellite mega-constellation,” Acta Astronautica, Vol. 202, 2023, pp. 653–669.

https://doi.org/10.1016/j.actaastro.2022.10.027.

24
 [91] Dishon, M., and Weiss, G. H., “A communications satellite replenishment policy,” Technometrics, Vol. 8, No. 3, 1966, pp.

399–409. https://doi.org/10.1080/00401706.1966.10490373.

[92] Kelley, C., and Dessouky, M., “Minimizing the cost of availability of coverage from a constellation of satellites: evaluation of

optimization methods,” Systems engineering, Vol. 7, No. 2, 2004, pp. 113–122. https://doi.org/10.1002/sys.10059.

[93] Zheng, H., and Ren, L., “Availability analysis of satellite constellation,” 2009 8th International Conference on Reliability,

Maintainability and Safety, 2009, pp. 245–248. https://doi.org/10.1109/ICRMS.2009.5270198.

[94] Jakob, P., Shimizu, S., Yoshikawa, S., and Ho, K., “Optimal satellite constellation spare strategy using multi-echelon inventory

control,” Journal of Spacecraft and Rockets, Vol. 56, 2019, pp. 1449–1461. https://doi.org/10.2514/1.A34387.

[95] de Weck, O. L., Scialom, U., and Siddiqi, A., “Optimal reconfiguration of satellite constellations with the auction algorithm,”

Acta Astronautica, Vol. 62, No. 2-3, 2008, pp. 112–130. https://doi.org/10.1016/j.actaastro.2007.02.008.

[96] Lee, H. W., and Ho, K., “Regional constellation reconfiguration problem: Integer linear programming formulation

and Lagrangian heuristic method,” Journal of Spacecraft and Rocets, Vol. 60, No. 6, 2023, pp. 1828–1845. https:

//doi.org/10.2514/1.A35685.

[97] Grogan, P. T., Golkar, A., Shirasaka, S., and De Weck, O. L., “Multi-stakeholder interactive simulation for federated satellite

systems,” 2014 IEEE Aerospace Conference, IEEE, 2014, pp. 1–15. https://doi.org/10.1109/AERO.2014.6836253.

[98] Grogan, P. T., and de Weck, O. L., “Interactive simulation games to assess federated satellite system concepts,” 2015 IEEE

Aerospace Conference, IEEE, 2015, pp. 1–13. https://doi.org/10.1109/AERO.2015.7119101.

[99] Grogan, P. T., Ho, K., Golkar, A., and De Weck, O. L., “Multi-actor value modeling for federated systems,” IEEE Systems

Journal, Vol. 12, No. 2, 2016, pp. 1193–1202. https://doi.org/10.1109/JSYST.2016.2626981.

[100] Guzzetti, D., Tauritz, D. R., Qureshi, R., Roberts, C., Indaco, M., Bone, L., and Kimbrel, E., “Satellite Tycoon: modeling

economic competition in the business of p-LEO constellations,” 11th International Workshop on Satellite and Constellations

Formation Flying, 2022.

[101] Qureshi, R., Roberts, C., Indaco, M., and Guzzetti, D., “Modeling and gamification framework of business competition

between p-LEO constellations,” Astrodynamics Specialist Conference, 2022.

[102] Qureshi, R. S., Roberts, C., Kimbrell, E., Mulder, S., Rao, A., Tauritz, D. R., and Guzzetti, D., “A table-top game to simulate

competition between p-LEO satellite internet constellations,” Astrodynamics Specialist Conference, 2023.

[103] Larson, R., and Odoni, A., Urban Operations Research, Prentice-Hall, New Jersey, 1984.

[104] Simchi-Levi, D., Chen, X., and Bramel, J., The Logic of Logistics: Theory, Algorithms, and Applications for Logistics

Management, Springer, 2014.

[105] Sztrik, J., “Finite-source queueing systems and their applications,” Formal Methods in Computing, Vol. 1, 2001, pp. 7–10.

25

View publication stats