Evaluation of Production Models in Large-Scale Satellite Manufacturing

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

th

5 Federated and Fractionated Satellite Systems Workshop


November 2-3, 2017, ISAE SUPAERO – Toulouse, France

Evaluation of production models in


large-scale satellite manufacturing
Gianluca Palermo, Paolo Gaudenzi
“La Sapienza” University, Rome, Italy
gianluca.palermo@uniroma1.it ; paolo.gaudenzi@uniroma1.it

Abstract

This paper proposes a method to assess how different production models can affect cost and value of a
large-scale production of satellites intended to be deployed in an Earth-orbiting constellation.

The study aims at finding the most appropriate production models for different satellite typologies and
constellation architectures. For example, whereas in the production of satellites of large size each
spacecraft subsystem must be assigned to a dedicated production unit, in the production of micro and
nano satellites such a scheme for the manufacturing process can be extremely inefficient or (in case of
highly-integrated designs) it can even make no sense.

The proposed method, on the basis of some key-characteristics of the spacecraft design and of the
constellation architecture, evaluates cost and value metrics for the production and testing activities, and
aims at identifying the most efficient trade-offs in the organization of the manufacturing process. Unit-
dedicated cellular workstations, and -at the opposite- a single product-assembly-line can be considered
the two extremes in the organization of the manufacturing process: in the middle of these two, a number
of crossbreed configurations can be considered to find the most appropriate production method according
to the intended satellite typology and the given constellation architecture.

Keywords

Satellite constellations; production lines; optimization

1
th
5 Federated and Fractionated Satellite Systems Workshop
November 2-3, 2017, ISAE SUPAERO – Toulouse, France

1. INTRODUCTION

In the past few years a number of new players have entered the satellite industry, proposing to launch
small and relatively inexpensive satellites whose capabilities are comparable to those offered by the
traditional large-size satellite typologies. This new trend in the space industry attracted multi-billion dollar
companies, such as Google and Facebook, who have shown interest in entering the market with the
intention of launching hundreds of small satellites to offer low-cost Earth Observation services or to
provide internet access to the developing countries [1][2].

The introduction of this new paradigm has started the competition with existing large-satellites
manufacturers, changing the rules of the game and posing a potential threat for the long-time industry
players, whose traditional production models start to be called into question. For this reason the adoption
of new models in the organization of the production activities is not only desirable in the present space-
industry scenario, but it is also essential for gaining efficiency and competitiveness.

In particular, the innovation in the organization of assembly lines, especially in the production -on a large
scale- of series of spacecraft destined to be deployed in an earth orbiting constellation, is a crucial aspect
for a successful business strategy [3].

Unfortunately, it is common practice in many industrial environments to disregard the holistic vision of the
problem. In fact, in many cases, a perception persists that the benefits obtained by the application of
advanced optimization techniques are not worth the additional effort required to change consolidated
schemes and procedures for the inclusion of such advanced methodologies [4][5].

The use of assembly lines dates back to the age of the Industrial Revolution during which the high
volumes reached in the mass production of cars made it necessary to find new ways of organizing the
production activities according to new rational criteria [6][7][8]. Since then, many advancements have
been made. Just in time, Lean manufacturing and Agile manufacturing models are only three examples of
innovations introduced in production schemes over time [1][5].

In the satellite industry, the adoption of advanced methods in the organization of the production flow is of
fundamental importance since the challenge of matching the delivery of new units with the launch
schedules and designated orbits is particularly complex [4][9]. The use of optimization methods -like
Pareto analysis- can facilitate the identification of those configurations for the production process that
result particularly suitable for a given constellation design and its associated deployment schedule [1][6].

Purpose of this paper is to propose a methodology to assist production-processes designers in the


identification of the best solutions for the configuration of the assembly lines.

2. ASSEMBLY LINES AND FLEXIBILITY CRITERIA


In order to better explain the rationale behind the proposed methodology we need to describe briefly how
an assembly line works. If we assume that the integration of a satellite is made up of seven stages, each
corresponding -for example- to a different subsystem being integrated on a dedicated workstation, then
the total integration time of a satellite corresponds to the sum of the single integration times of each
stage. Anyway by pipelining the production of the subsequent satellites in a way that the following unit
enters a workstation as soon as the workstation becomes available, the assembly line can deliver a
finished satellite with a frequency equal to the one of the longest-lasting production stage instead of a
frequency equal to the total integration time of a satellite, resulting in a considerable advantage.

2
5th Federated and Fractionated Satellite Systems Workshop
November 2-3, 2017, ISAE SUPAERO – Toulouse, France

Thus, the longest lasting stage determines the pace of production and represents a bottleneck which can
be the object of optimization. In general the longest lasting stages are the targets of the optimization
process because intervening on them allows to reduce the output frequency of the assembly line [7][8].

Figure 1 A typical serial assembly line

Figure 1 represents a typical assembly line in which stage integration time (on-station dwelling time),
workstation hardware cost, and station-related labor cost-per-hour are indicated for each stage. Values
are represented in conventional units for illustrative purposes.

One simple way of dealing with the bottleneck, represented by stage B in Figure 1, is the parallelization of
stage B in two or more threads [6]. Figure 2 shows an example in which stage B is parallelized in two
threads: this modification allows to halve the average dwell time of a unit in stage B.

Figure 2 A serial assembly line with a parallelized stage

Another, more complex, way of intervening on the assembly line configuration is the merger of adjacent
stages into a single stage. This is shown in Figure 3 with the merger of B and C stages. In the same
Figure, the merger of E and F stages is shown as well.

Figure 3 A serial assembly line with merged stages

This merger can be done when similar activities are to be carried out: for example the TT&C subsystem
and a Telecommunication payload could be integrated on the same workstation, taking advantage of the
fact that both subsystems require assembly facilities pertinent to electronics. Small, highly integrated,

3
5th Federated and Fractionated Satellite Systems Workshop
November 2-3, 2017, ISAE SUPAERO – Toulouse, France

satellites represent an example of a product that would benefit most from such a modification in the
production chain.
The effects of a merger of two or more production stages are not so simple to evaluate as it is in the case
of the parallelization of a production stage.
For this reason a preliminary analysis has to be conducted to determine how some key parameters (like
on-station dwelling time, workstation hardware cost and station-related labor cost-per-hour) are affected
by the merger.

Typically the effects of a merger can be summarized in 3 main aspects (cf. example in Figure 4)
[10][11][12]:

• A reduction (w.r.t. the sum of stages’ values, indicated by the green marker in Figure 3) of the dwell
time (t) deriving from the possibility to perform simultaneously some of the tasks;
• An increase (w.r.t. the sum of stages’ values, indicated by the red marker in Figure 3) of the labor
cost per hour (lc): this is due to a higher degree of specialization needed for operators which are
required to deal with all the tasks pertinent to more than one production stage at the same time
(operator capabilities are no longer limited only to a single production stage but they must be able to
deal simultaneously with the tasks deriving from all the original merging stages).
• A possible increase or decrease (w.r.t. the sum of stages’ values, indicated in Figure 3 by the green
and red markers, respectively) of the workstation hardware costs (sc) according to the degree of
versatility of the tools provided with the workstation itself and to the synergies that emerge within.

Figure 4 Examples of labor, cost and time modifiers that apply to the merger of two different couples of production
stages (B+C and E+F stages). Opposite arrows in the “workstation merger cost modifier” column indicate that the
merger operation does not result in an univocal trend for such modifier.

A combination of parallelization and merger solutions is also possible, as shown in Figure 5, giving
maximum flexibility when searching for a solution to the problem of the identification of an optimal
assembly-line configuration.

Figure 5 A serial assembly line with a merged stage being parallelized

In the next section, a general methodology will be proposed for the generation of a number of possible
parallelization and merger alternative solutions and for the identification of the best assembly-line
configurations according to cost and time metrics.
4
5th Federated and Fractionated Satellite Systems Workshop
November 2-3, 2017, ISAE SUPAERO – Toulouse, France

3. A METHODOLOGY FOR THE OPTIMAL CONFIGURATION OF ASSEMBLY


LINES

The proposed methodology is made up of 14 steps and is presented in detail in the functional flow block
diagram of Figure 6.

Figure 6 Functional flow block diagram of the proposed methodology for the evaluation of different assembly line
configurations and the identification of optimal ones

5
5th Federated and Fractionated Satellite Systems Workshop
November 2-3, 2017, ISAE SUPAERO – Toulouse, France

The rationale of the methodology is to harmonize the pace of production with the pace of deployment
phases: this is done by acting on parallelization and merger of production stages in order to tune up the
assembly line output rate with the deployment plan and launch schedule [9].

The methodology starts with the definition of the constellation configuration and of the in-orbit deployment
plan.

It assumes the preliminary knowledge, for each stage, of all the parameters of the basic (linear) assembly
line, purposely the aforementioned t, lc, sc.

The methodology requires the identification of which production stages can be parallelized (and at which
extent), together with the identification of which production stages can be merged including the
specification of the new values for the parameters t, lc, sc.

It proceeds with the generation of a number of valid alternative stage-configurations (within the limits
defined in the previous steps for parallelizations and mergers), as detailed in Figure 7 (this activity is part
of step 6 of Figure 6).

The enumeration of the configurations given by all the possible combinations of the generated
alternatives is performed afterwards.

Figure 7 Algorithm for the automatic generation of valid alternative configurations through the generation of various
parallelization and merging combinations
6
th
5 Federated and Fractionated Satellite Systems Workshop
November 2-3, 2017, ISAE SUPAERO – Toulouse, France

Cost and time, which represent the two performance evaluation metrics, and whose definition is given in
Table 1, are then calculated for each configuration and confronted against launch plans [9]. Economies of
scale and discount rates calculations are also considered in the cost evaluation.

Table 1 Definition of evaluation metrics and the breakdown of the respective component items

Metrics Components

Time Total production time Launch wait time - -

Production variable Missed


Cost Production fixed costs Launch
costs opportunities

Finally a Pareto downselection criterion is applied to identify the most efficient assembly-line
configurations that comply with the deployment and launch schedule requirements [13][14][15].

Through the proposed methodology production-process designers can better comprehend assembly-line
configuration trade-offs and make informed decisions on how to modify existing processes, or develop
brand-new ones.

4. CONCLUSIONS

This paper proposes a methodology aimed at assessing how different production models can affect cost
and value of a large-scale production of satellites intended to be deployed in an Earth-orbiting
constellation.

The rationale of the methodology resides in the attempt to harmonize the pace of production with the
pace of deployment phases: this is done by acting on parallelization and merger of production stages in
order to tune up the assembly line output rate with the deployment plan and launch schedule.

Different combinations of parallelization and merger of production stages are generated, confronted with
deployment plans and launch schedules, and evaluated through cost and time metrics.

Efficient tradeoffs are identified through the application of a Pareto downselection criterion.

The proposed methodology is intended to assist production-processes designers in the identification of


the best solutions for the configuration of the assembly lines in satellite mass productions.

5. REFERENCES
[1] Carayannis E.G., Roy R.I.S. The speed and acceleration of technological innovation in the small satellite
manufacturing industry: a co-opetitive dynamics perspective, International Conference on Engineering and
Technology Management, 1998.
[2] K. Maine, C. Devieux, P Swan, Overview of Iridium Satellite Network, Microelectronics Communications
Technology Producing Quality Products Mobile and Portable Power Emerging Technologies, 1995.

7
th
5 Federated and Fractionated Satellite Systems Workshop
November 2-3, 2017, ISAE SUPAERO – Toulouse, France

[3] J.H. Saleh, D.E. Hastings, D.J. Newman. Extracting the essence of flexibility in system design. Proceedings Third
NASA/DoD Workshop on Evolvable Hardware. EH-2001, 59-72.
[4] Wilson, J. M., & McKinlay, A.. Rethinking the assembly line: Organisation, performance and productivity in Ford
Motor Company, c. 1908-27, 2010
[5] T. Taylor. “Evaluating and Selecting Manufacturing Flexibility”. S.M. thesis. Department of Mechanical
Engineering, Massachusetts Institute of Technology, 1991.
[6] Wilson, J.M., Henry Ford vs. assembly line balancing. International Journal of Production Research, 2014.
[7] Ford, Henry & Crowther, Samuel, My Life and Work. Garden City, NY: Garden City Publishing, 1922.
[8] Hounshell, David A., From the American System to Mass Production, 1800-1932: The Development of
Manufacturing Technology in the United States. Baltimore: Johns Hopkins University Press, 1984.
[9] J. Frank, A. Jonsson, R. Morris, D.E. Smith, Planning and scheduling for fleets of Earth observing satellites, in:
Proc. 6th International Symposium on A.I., Robotics and Automation for Space, 2001.
[10] T. C. Hu, “Parallel Sequencing and Assembly Line Problems”, Operations Research 9 (6), 1961.
[11] I. Baybars, “A Survey of Exact Algorithms for the Simple Assembly Line Balancing Problem,” Management
Science, Vol. 32, No. 8, 1986.
[12] Becker C., Scholl A. A survey on problems and methods in generalized assembly line balancing. European
Journal of Operational Research 168 (3), 2006.
[13] O.L. de Weck, 16.842 Fundamentals of systems engineering — Lecture notes, MIT OpenCourseWare,
Cambridge, MA, 2009.
[14] O.L. de Weck, M.B. Jones, Isoperformance: Analysis and Design of Complex Systems with Desired Outcomes,
Systems Engineering, 1 (2006) 45-61.
[15] A Salado, Defining Better Test Strategies with Tradespace Exploration Techniques and Pareto Fronts:
Application in an Industrial Project, Systems Engineering 18 (6), 639-658, 2015.

You might also like