Conference PaperPDF Available

Autonomous Additive Construction on Mars

Authors:
  • Foster + Partners
Autonomous Additive Construction on Mars
Samuel Wilkinson
1
, Josef Musil
2
, Jan Dierckx
3
,
Irene Gallou
4
and Xavier de Kestelier
5
Specialist Modelling Group, Foster and Partners,
Riverside, 22 Hester Road, London, SW11 4AN, UK.
1
swilkinson@,
2
jmusil@,
3
jdierckx@,
4
igallou@,
5
xdekeste@fosterandpartners.com
ABSTRACT
We present a conceptual construction process for an inhabitable outpost on Mars, using
an autonomous multi-robot swarm approach to additively sinter layers of regolith into a
protective shield over an inflatable pressurised module. The guiding design hypothesis
investigated is that physically distributing risk across multiple simpler units working in
parallel can improve chances of success, rather than the traditional consolidation into
a single complex unit. This approach is fundamentally enabled by the decreasing size
and cost of hardware. Larger numbers of more intelligent robots offers the possibil-
ity of emergent behaviour as the collective action of a complex group can be greater
than that of individual independent units. In the paper we consider the general bene-
fits of distributed redundancy, its implications for a robotic construction process using
microwave power to selectively sinter layers of regolith in-situ, a subsequent habitat de-
sign with regolith shield, and a roadmap for a future technology demonstration.
INTRODUCTION
As various private and public organisations propose numerous missions to the moon,
near-Earth asteroids, or Mars in the coming decades, certain trends in current solutions
to surface habitation are observable. These include the use of: i) an initial robotic
expeditionary mission to prepare for subsequent human arrival; ii) the combination
of a delivered pressurised module for habitation and an in-situ compressive shield of
regolith; and iii) large-scale additive construction with regolith. Assuming the first two
points, and although there is of course dependence between them, we focus on the
third for which there is specific debate about the best additive construction process.
By construction process, it refers to both: the robotic system, i.e. a static vs. mobile
configuration, the scale, capability, and number of units; and secondly, the material and
bonding method.
Additive manufacturing offers many possibilities when applied in this context at larger
scales to the construction of extra-planetary human habitats. If coupled with a robotic
system, it enables the use, prior to human arrival, of local in-situ materials to work
in conjunction with other delivered habitat elements. There are obvious complimen-
tary advantages in delivering inflatable light-weight pressurised modules and a system
to construct a heavy compressive regolith shield on top. The light-weight inhabitable
element offers the high-tech aspects of life-support, manufactured in controlled envi-
ronments on Earth, whilst the heavy-weight shield acts as protection from radiation,
micro-meteorites, and dust storms.
In the following sections we firstly describe the concept of distributed vs. consolidated
redundancy which we posit will be possible in multi-robot systems in the near future.
In the second section, we introduce the concept of a multi-robot swarm and its theo-
retical emergent behaviour. We then give a taxonometric overview of existing additive
construction technologies, identify those with possible application, and introduce the
process of melting regolith layers via microwaves into a proposed habitat shield with
designs for three classes of construction robot. These sections come together into a
proposed further work technology demonstration.
REDUNDANCY
System redundancy is often achieved through either passive or active means, where
passive redundancy means an added strength capable of absorbing extra strain (this
is common practice in structural engineering), and active redundancy where separate
backup systems are present to come online when primary ones fail (common in elec-
tronic and computer engineering). In active redundancy, the extra backup capabilities
are typically isolated or on standby, yet still integrated into the main system. In a dan-
gerous environment this presents an issue when we consider that they are physically
situated in a single fate-sharing machine; all our eggs are in one expensive basket. For
example, a satellite may have passive redundancy in its physical build and active re-
dundancy in its multiple spare computers on-board, yet its success is under risk due to
consolidation of this complexity into a single device. An alternative paradigm would
envisage distribution of the functionality, and risk, across multiple devices which can
act together for the mission.
Many, if not all, complex systems in biology have a ‘fine-grained’ architecture (Mitchell,
2009), in that they consist of large numbers of relatively simple elements that work to-
gether in a highly parallel fashion. This fine-grained nature of the system not only
allows many different paths to be explored, but it also allows the system to continu-
ally change its exploration paths, since only relatively simple micro-actions are taken
at any time. Employing more coarse-grained actions (for instance with a single robot)
would involve committing time to a particular exploration that might turn out not to
be warranted. In this way, the fine-grained nature of exploration allows the system to
fluidly and continuously adapt its exploration as a result of the information it obtains.
Moreover, the redundancy inherent in fine-grained systems allows the system to work
well even when the individual components are not perfectly reliable and the informa-
tion available is incomplete or local. Redundancy allows many independent samples of
information to be made, and allows fine-grained actions to be consequential only when
taken by large numbers of components.
By the law of accelerating returns (Kurzweil, 1999, 2004), a generalisation extend-
ing Moore’s law to technological development, the rate is observably exponential, i.e.
whilst the cost and size of computation tends to zero, performance increases to infinity.
Assuming this is broadly true, as it appears to be, and that it continues into the imme-
diate future, we can take the premise that ever-increasing functionality will be possible
on ever-smaller, ever-cheaper machines. This will in turn enable progressively smaller
and capable robots suitable for enabling our own presence in space. We therefore posit
that it is both beneficial and possible in the near future to distribute risk across multiple
such robots to construct ourselves a habitat on another planet.
SWARM ROBOTICS
In the near future it is predicted that computational intelligence (CI) and robotic tech-
nology will be sufficiently advanced to allow for a distributed system of autonomous in-
telligent machines. A system capable of adapting to uncertain operating environments,
of self-management and system awareness, and of following high-level commands such
as ‘explore’, ‘gather materials’ or ‘construct habitat’ (Nilsson, 2014; Truszkowski et al.,
2006).
Behaviour has significance for understanding biological intelligence. It is understood
as interactions between an organism and its environment where the actions of the organ-
ism affect its own perceptions, and thus its future actions and perceptions. Applying
this concept to robots gives the field of autonomous robots that are behavioural ma-
chines capable of operating in partially unknown and changing environments without
human intervention (Floreano and Mattiussi, 2008).
We apply these principles of behaviour and self-organization to collections of simple,
autonomous robots. Simple is understood as not having sophisticated sensors, elec-
tronics or mechanics. Neither do they use global information to be centrally controlled
and as a result algorithms tend to be simpler. The core idea of swarm robotics is to
capitalize on simple interactions among robots in order to solve complex problems by
means of emergent behaviour. This concept is inspired by a similar phenomenon found
in nature, for example social insects.
Swarm robotics is mainly concerned with groups of robots that are larger than groups
easily controlled with a centralized, top-down approach but within a number that is
manufactured at the same or lower cost of a few complex robots. The main poten-
tial advantages of this approach is the robustness of the swarm to failure of individual
robotic units. Further advantages are addition of units in real time and without chang-
ing the emergent behaviour, ability to cope with environmental noise and individual
differences, and the emergent effect where the work of the swarm is greater than the
sum of the work by the individual units.
Swarm robotics and uncertainty
Probabilistic robotics was developed after the 1970s when most research in robotics
presupposed the availability of exact models of robots and their environments (Thrun
et al., 2005). Little emphasis was placed on sensing and the intrinsic limitations of
modelling complex environments. In the mid 1980s the paradigm shifted towards reac-
tive techniques. Reactive controllers rely on capable sensors to generate robot control.
Rejections of models were typical for researchers in this field. Since mid-1990s a new
approach has begun to emerge: probabilistic robotics. This approach relies on statisti-
cal techniques to seamlessly integrate imperfect models and imperfect sensing.
The Martian environment is computationally difficult to simulate. The advantage of
cooperative robots is that they operate locally within the immediate environment only
and they do not need to understand the whole complexity, thus it is easier to simulate.
Robots can use sensors to measure what has been built and update a digital model of the
next printed layer to compensate for tolerances. Decentralisation provides a great ro-
bustness as no single failure can cause an overall failure (Brambilla et al., 2013).
AUTONOMOUS ADDITIVE REGOLITH CONSTRUCTION
Additive Construction (AC) refers to the application of Additive Manufacturing (AM)
to an architectural scale; that is, the 3D printing of buildings. The distinctions are
therefore application and scale, with AM typically for pieces less than 1m and AC
for anything greater. Larger scale AC necessitates a higher deposition rate than AM,
a larger construction system due to the greater size of prints, and usage of different
materials. Each of these has a secondary effect of reducing the print resolution or
accuracy to on average the order of 10mm, whereas AM today achieves sub-millimetre
resolution.
In developing a taxonomy of additive construction techniques (Table 1), all examples
can firstly be classified by one of two deposition methods: either process material in-
ternally then deposit (e.g. AM fused-deposition modelling (FDM)); or deposit then
process externally (e.g. powder-based AM 3D printing). We consider the latter to
be more practical in uncontrolled environments with variable material sources since it
does not require internal processing which may lead to blockages.
The second classification is the binding method, generally divided into bonding (via an
adhesive or chemical reaction) or by heating (sintering or melting). Sintering means
heating until individual particles fuse together without melting to liquefaction (Figure
4). We propose that sintering consecutive layers of materials together is more logical as
it is less energy intensive than completely melting and does not require transportation
or in-situ processing of adhesives.
The third category is the material used for construction. In the examples given, this
includes plastics, glasses, ceramics, metal, sand, concrete, and regolith. Without more
complex in-situ material processing technologies, we are essentially left with three
viable methods of layered in-situ regolith sintering: microwave, laser, or solar sinter-
ing.
In Table 1, the examples of AC highlighted in light grey have all used real or simulant
lunar or Martian regolith as construction material. These have all been considered for
or have potential for use in our context. Those highlighted in dark grey use regolith
and layered in-situ microwave sintering, and are the closest precedent technology. All
of the examples given have been demonstrated to some extent on Earth.
Site preparation
Before construction can begin however, the system must be delivered to the surface
and the site must be prepared. The Entry, Descent and Landing (EDL) will happen
Table 1: Taxonometric overview of additive construction technologies
Construction
Method
Binding Method Base Material Additives
Support
Material
Comments Example
YES
Portland/Sorel cement are not
appropriate in vacuum. Regolith + sulfur
is an ISRU option.
Loughborough University,
Skanska, F+P et al.
WinSun
Bruil
WASP
BetAbram
CyBe
Spetsavia
xTreeE
Le Roux
TU Eindhoven
Con3D
Rudenko
Contour Crafting USC
Dirk van der Kooij
TechmerES, ORNL, SOM
KamerMaker
Glass - - - MIT Mediated Matter
IAAC
Robocasting SNL
Microwave melting Regolith - - Melted in chamber and extruded. JPL, PISCES
Gluing Regolith Urethane - Not appropriate for large scales. Adherent
Powder spray Regolith / Sand Water, air -
In a vacuum only possible if molten
powder is sprayed.
IAAC
Norsk Titanium
Cranfield University
D-shape
Voxeljet
Regolith - YES - Loughborough University
KSC
PISCES
- - - Aachen Uni
-- - NUS
- YES - EOS
Sand - YES - Kayser, M. (MIT)
NASA KSC
PSI
JPL, PISCES
University of Knoxville
Selective inhibition
sintering
Regolith
With or Without
(MgO, Portland
C/Water)
YES
Requires pressure. Also used for
additive assembly e.g. tiles.
USC
Melted in-situ. Good penetration
properties. Can be assisted by infrared
heating. Difficult to shape.
Laser sintering
Basalt - -
Microwave sintering Regolith - -
Regolith - - -
Time consuming, energy intensive.
Extrusion
deposition
-
Advantageous in vacuum.
Regolith / Sand
Layered in-situ
Chemical Regolith / Sand MgCl
2 YES
Solar sintering
Portland/Sorel cement are not
appropriate in vacuum. Regolith + sulfur
is an ISRU option.
Fused-deposition
method (FDM)
Plastics - - -
Ceramics
Chemical Sand
Sulfur, Portland
cement, plastics,
Sorel cement
-
--
Requires pre-processed metal, time
consuming.
Additive welding /
Direct metal deposition
Metal - -
in two phases: first the robots will be delivered to the surface for site selection and
preparation; followed later by the ø4.6m habitat modules. For each, once stationary
on the surface, the individual modules may be distributed within a landing radius of a
few hundred metres, therefore it is necessary for them to navigate together and collect
at a designated common point. A sequenced inflation-deflation of the external air bags
starts a controlled roll of the modules to a shared target.
The first task for the robot system once deployed at the site is to excavate a 1.5m-deep
hole for the habitat modules to sit within. The largest class of robots (RAC-D) will
dig the loose regolith from the surface layer by layer, which will be moved nearby into
protective berms by the medium-sized (RAC-T) robots (Figure 1a). The volume of ex-
cavated regolith is roughly equal to the amount to be printed on the shield. There may
be an element of trial-and-error in the initial excavation due to unknown underground
geology (i.e. hidden rocks), although this could be minimised by either selecting ap-
propriate land formations such as a small crater, or by acoustic location.
(a) Multi-robot EDL and site preparation (b) EDL and navigation of habitat units
(c) Stages of module deployment: (right to left) opening, inflation, and connection
Figure 1: Initial site preparation
Once the site is prepared, and the three modules are gathered together (Figure 1b), the
small inflatable spheres surrounding the habitat partially deflate. The upper faces of
the dodecahedron module fold up and lock in position, similarly for the lower faces to
form the foundation which can fit to a rough landscape for the interior floor to be level.
Subsequently, the core, shaped as a pentagonal prism, expands outwards. Each of the
ve vertical faces of the core is either a connection or an airlock which will move to
the outer perimeter along with the inflated skin. The three deployed habitat modules
are now ready for the regolith shield to be constructed on top (Figure 1c).
Regolith construction
The regolith additive construction (RAC) approach is designed for the low accuracy
likely to be expected from using variable materials at an uncertain site with autonomous
robots in the field. The reliability is partly in its distribution of tasks, implemented by
the three classes of robot: the strategy is to ’dig, move, and melt’ regolith.
The large digging robots will extract loose regolith in close proximity for the medium-
sized mover robots to transport to the habitat. As the regolith is deposited, the smallest
melting robots have a microwave print head to bond one layer at a time (Barmatz et al.,
2014). The regolith is positioned into rough layers of about 10mm thickness by the
transporter robots, with the thickness continuously measured (Figure 3. Once a thin
layer of regolith is in place, the third class of smallest robots selectively sinters patches
into a hard crystalline material. The form and progressive construction of the shield is
shown in Figure 2.
(a) Pre-construction (b) Back-filled (c) 25%
(d) 50% (e) 75% (f) Complete
Figure 2: Plan view of progressive regolith construction
Since each of the three modules has five possible connections, three of which are for
entry, these areas must be left clear. As the construction increases in height, the perpen-
dicular ramps extending radially allow for the robots to progress to the next layer.
The form of the regolith shield is driven by two key criteria. The first criteria is the min-
imal thickness of regolith needed to protect the inhabitants from radiation: a smaller
one is applied to two of the three modules and the larger value for the third remaining
module. This means that one module can serve as a safe space with extra protection
in case of transient solar flares. For protection from radiation over long-term periods,
rather than transporting heavy shielding from Earth, the regolith shell is a logical al-
ternative. The largest reduction in dose equivalent (rem) occurs in the first 20g/cm
2
,
so assuming a regolith density of 1.5g/cm
3
the regolith depth should be at least 15cm
(Simonsen and Nealy, 1991). Whilst this is a minimum depth to ensure the survival
of the inhabitants, the design includes 1.5m above the work/sleep modules, and 2.5m
above the communal space. This will improve the long-term health of the astronauts
and provide a temporary shelter during periods of increased solar activity.
(a) Deposition and sintering of regolith layers
by multiple RAC-T and RAC-M robots
(b) Operational occupied outpost with contin-
uous repair and infrastructure construction
Figure 3: Additive construction process
The second factor of the shield’s form is the ability of all construction robots to transfer
themselves to the highest layer printed so far during the construction process. As a
result, multiple ramp structures blended into the overall form are introduced next to
every opening of each module (airlocks, windows and suitports). Because of their
location, they also serve as an extra protection of these openings.
Microwave sintering
There are various possible methods available for bonding loose regolith for construc-
tion: namely chemical; with adhesive; freezing; compaction; laser or solar sintering; or
microwave sintering. The last is investigated here due to its apparent robust application.
Taylor and Meek (2005) and Barmatz et al. (2014) have both attempted sintering of real
Lunar regolith and Martian regolith simulant (JSC-2A) respectively in the laboratory
with 2.45GHz 200W microwaves. The bonding mechanism of sintering lunar regolith
is shown in Figures 4a to 4d, and an example of the internal structure of a bonded
sample in Figure 4e. Due to the thermal properties of the material, the sample has a
heterogeneous layering of loose material, to partially bonded, through to a completely
molten core.
(a) (b) (c) (d)
Loose
powder
Sintered rind
Melted interior
(smooth, glassy)
(e)
Figure 4: Regolith sintering: (a-d) Progressive sintering of lunar regolith (Hintze et al.,
2009); (b) Interior of lunar JSC-2A after being heated to ˜600
C (Barmatz et al., 2014)
As classified in the earlier taxonomy, there are two methods to applying this either
internally or externally. The first involves placing a layer of material and sintering
it in-situ Taylor and Meek (2005). For example, the ‘lunar road-paving wagon’ can
move back and forth with its magnetrons (microwave generators) that can be set to
various frequencies and power, in order to effectively sinter the lunar regolith, thereby
constructing a traffic-able road or launchpad.
(a) (b)
Figure 5: Regolith sintering in-situ: (a) ‘Lunar Road-Paving Wagon’ designed by Tay-
lor and Meek (2005); (b) Microwave printer concept by Barmatz et al. (2014)
In the second case, regolith is collected into a reservoir and fed through a microwave
oven where it becomes molten and is subsequently extruded out into position (Barmatz
et al., 2014).
Power beaming
The robot’s batteries charge wirelessly via power beaming (Brown, 1996), a technol-
ogy that is currently under development. This involves transmitting power over long
distances through air, space, or optical fibres via a laser and receiving photovoltaic
cells. Similar to solar power although much more intense, the transmitting station con-
verts electricity to laser light which is transmitted to photovoltaic cells on the individual
robots, where it is finally converted back to electricity.
The great advantage of wireless power is in enabling ‘perpetual presence’, i.e. with a
constant supply of power the robots do not have to rely on finite battery lives. Tech
demos with a quadcopter have achieved over 12 hours of constant flight (Nugent and
Kare, 2010).
Swarm robotics behaviour
Distributed system of powering also allows for higher number of smaller robots to
be controlled operate independently. A task harder to realize with larger robots, where
each needs to carry its own power source able to power it over long period of time.
A programmable framework for a cooperative swarm robotic system that builds large
scale architectural objects by means of additive manufacturing is here investigated.
Greater precision material deposition is further enhanced by on-site measurements to
compensate for uncertainty in its simulation having stochastic parameters. Additive
manufacturing is nowadays well established for small scale prototyping and manufac-
turing but needs to scale up in architectural context. Current research focuses on printed
material research and leaves issues with architectural scale to either conventional as-
sembly or limiting the overall size of printed object. Cooperative swarm robotics is
a distributed system that allows for large scalability. Real time physical measure-
ments are proposed in-conjunction with virtual simulation to react to under-defined
and changing environmental noise.
Swarm robotics at Harvard (Werfel et al., 2014, 2011) shows a multi-agent construction
system inspired by mound-building termites. First case of study tries to reverse engi-
neer how to build a given form by a given number of independent robots. That solves
the inverse problem of how low-level rules give specific outcomes, that is in general
still little understood. Suggested is a further development of stochastic rules for agents
for not fully pre-determined forms. This should be further developed and understood in
the near future and will be suitable for distant places like Mars, where precise delivery
of predefined from is not necessary, rather delivery of precise properties, like isolation
and thickness of walls is necessary.
Another advantage of swarm robotics is the emergent outcome. Kilobot project at
Harvard (Rubenstein et al., 2012) studies this emergent effect while keeping the price
and complexity of each robot very low.
Robot design
The multi-robot system consists of three classes of robot deployed for site preparation
and construction. For system diversity, robot movement is by wheels, tracks and legs
respectively (Figure 6). Each robot has the ability to operate independently without
any central command; therefore they need the ability for environmental sensing, local
communication, and decision making. Any expected emergent behaviour from the
group is dependent largely on the interactions between the individuals, either directly
via communication or indirectly through the environment which they alter.
(a)
1m 0.4m 0.25m
(b)
Figure 6: Robot design: (large to small) RAC-D, RAC-T, RAC-M
Table 1: Construction robot specifications
Name Function Size Quantity Movement
RAC-D ‘Digger’, regolith excavation us-
ing a perpendicular bucket-wheel
1.00m 1 Wheels
RAC-T ‘Transporter’, moves and deposits
regolith into thin layers
0.40m 5 Tracks
RAC-M ‘Melter’, microwaves patches of
regolith in desired position
0.25m 10 Legs
FURTHER WORK: TECH DEMO
The next developmental steps in implementing this proposal can be broken down into
four strategic development areas to be pursued. Firstly, operation of the robot system:
development of individual robot mechanics, control, and communication; group test
with multiple units to investigate the inverse problem of guaranteeing a satisfactory
end result with input rules (physical and simulation); and power beaming from cen-
tral power station to individual robots. Secondly, the additive construction process:
material collection, processing, and layering; microwave sintering of in-situ regolith
simulant; and combination of robotic and AC processes. Thirdly, habitat module de-
sign: EDL, navigation, inflation, deployment and connection; full-scale mock-up test
with human test subjects. And finally, system integration with a fully automated field
test in analogous Earth site.
CONCLUSION
The various aspects of redundancy that are explored here with the outpost delivery,
construction, and design are intended to increase chances of success. A key element
of this approach is allowing for flexible outcomes: from the initial navigation of the
modules to find a suitable location, with the internal module layout, through to the
semi-autonomous robotic additive construction.
In a largely uncharted extreme environment such as Mars, there are strict requirements
on and of any such construction system. However, its specific activities and subsequent
products must be less strictly controlled. Due to communication challenges and envi-
ronmental unknowns, complete direct control is not possible and therefore the auton-
omy of, or trust placed in, the system must be high. In truth this is the case for robotic
and future human missions. Relinquishing total control whilst ensuring through indi-
rect methods the desired outcomes is the greatest challenge of this endeavour.
Acknowledgements
This work was initially entered in the AmericaMakes 3D Printed Habitat design com-
petition, one of NASAs Centennial Challenges. We would like to thank the follow-
ing companies and individuals for their assistance during the competition: Astrobotic
Technology, Penelope Boston of New Mexico Tech, Malika Beggour, John Eager of
the British Antarctic Survey, David A. Green of King’s College London, and Rapha
Clothing.
References
Barmatz, M., Steinfeld, D., Anderson, M. and Winterhalter, D. (2014), 3D microwave
print head approach for processing lunar and mars regolith, in ‘Lunar and Planetary
Science Conference’, Vol. 45, p. 1137.
Brambilla, M., Ferrante, E., Birattari, M. and Dorigo, M. (2013), ‘Swarm robotics: a
review from the swarm engineering perspective’, Swarm Intelligence 7(1), 1–41.
Brown, W. (1996), ‘The history of wireless power transmission’, Solar energy 56(1).
Floreano, D. and Mattiussi, C. (2008), Bio-inspired artificial intelligence: theories,
methods, and technologies, MIT press.
Hintze, P. E., Curran, J. and Back, T. (2009), Lunar surface stabilization via sintering
or the use of heat cured polymers, in ‘47th AIAA Aerospace Science Meeting’.
Kurzweil, R. (1999), The age of spiritual machines: how we will live, work and think
in the new age of intelligent machines, Orion.
Kurzweil, R. (2004), The law of accelerating returns, Springer.
Mitchell, M. (2009), Complexity: A guided tour, Oxford University Press.
Nilsson, N. (2014), Principles of artificial intelligence, Morgan Kaufmann.
Nugent, T. and Kare, J. (2010), ‘Laser power for UAVs’, Laser Motive White Paper-
Power Beaming for UAVs, NWEN .
Rubenstein, M., Ahler, C. and Nagpal, R. (2012), ‘Kilobot: A low cost scalable robot
system for collective behaviors’, 2012 IEEE International Conference on Robotics
and Automation .
Simonsen, L. and Nealy, J. (1991), Radiation protection for human missions to the
moon and mars, Technical report, National Aeronautics and Space Administration,
Hampton, VA (USA). Langley Research Center.
Taylor, L. and Meek, T. (2005), ‘Microwave sintering of lunar soil: properties, theory,
and practice’, Journal of Aerospace Engineering .
Thrun, S., Burgard, W. and Fox, D. (2005), Probabilistic robotics, MIT press.
Truszkowski, W., Hinchey, M., Rash, J., Rouff, C. et al. (2006), Autonomous and
autonomic systems: A paradigm for future space exploration missions’, Systems,
Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on
36(3), 279–291.
Werfel, J., Petersen, K. and Nagpal, R. (2011), Distributed multi-robot algorithms for
the termes 3d collective construction system, Institute of Electrical and Electronics
Engineers.
Werfel, J., Petersen, K. and Nagpal, R. (2014), ‘Designing collective behavior in a
termite-inspired robot construction team’, Science 343(6172), 754–758.
... Today, we believe there is a strong and clear need for a fresh civil engineering vision, following upon the novel architectural propositions of De Kastellier et al. (2015) and Wilkinson et al. (2016). More specifically, there is a need for regolith-based ET structures that will exhibit resilience against natural hazards, also considering dynamic loading in the form of seismic ground motions and impacts, and constructed using large-scale additive manufacturing, interlocking regolith bricks, or other compaction/sintering techniques. ...
... Figure 36. a) Entry, descent and landing (EDL) of multi-robots for site preparation; b) EDL and navigation of the habitat units; c) deployment of modules (opening, inflation and connection),Wilkinson et al. (2016). ...
... View progress of regolith construction,Wilkinson et al. (2016). ...
Article
The design of a permanent human habitat on a planetary body other than the Earth is an idea introduced many decades ago, which became even more significant after the landing of the first humans on the Moon with the Apollo missions. Today's rampant technological advances combined with ambitious missions, such as the Insight mission on Mars and the Artemis program for the Moon, render the vision of space colonization more realistic than ever, as it constantly gains momentum. There is a considerable number of publications across several disciplines pertaining to the exploration of Lunar and Martian environments, to those planets' soil properties, and to the design of the first habitable modules. The scope of this paper is to present a meticulous selection of the most significant publications within the scientific areas related to: (a) geotechnical engineering aspects, including the mechanical properties and chemical composition of Lunar and Martian regolith samples and simulants, along with elements of anchoring and rigid pads as potential forms of foundation; (b) ground motions generated by different types of Moonquakes and meteoroid impacts; (c) the different concepts and types of extraterrestrial (ET) structures (generic, inflatable, deployable, 3D-printed), as well as overall views of proposed ET habitats. Apart from the details given in the main text of this paper, a targeted effort was made to summarize and compile most of this information in representative tables and present it in chronological order, so as to showcase the evolution of human thinking as regards ET structures.
... The Smartgeometry [17] 2016 workshop originated from the design concept for a habitat on Mars [23,24] ( Figure 1) in which inflatable modules are covered by a protective regolith shield. The primary component of the concept design that we consider is security, in terms of both crew safety, peace of mind, and mission success in relation to the environmental conditions. ...
... Additionally, protective structures can be built, such as berms, and infrastructure like landing pads and roads. [23,24]. ...
... This is the element of the construction process that is investigated further in the following workshop description. Preliminary site preparation, module deployment, inflation, and connection [23,24]. ...
Conference Paper
Full-text available
We present our findings from a 4-day workshop at SmartGeometry2016, during which the authors conducted an open-ended experiment to ascertain the viability of a multi-robotic system capable of large-scale additive construction with sand. The study and results pertain to future robotic systems operating in extreme environments, such as in lunar or Martian conditions, where there is a need for autonomous construction of regolith structures for infrastructure or human habitat protection. The purposes of this study were to: i) implement and document the practical knowledge of multi-or swarm-robot systems in physically realistic environments; ii) ascertain the feasibility of additive construction with many-simple rather than few-complex robots; and iii) explore individual behavioural rules for the robots which, although indirectly controlled themselves, can result in a controlled outcome. Behaviour is understood as the interactions between an individual and its environment where the behaviour of the individual affects its own perceptions, and thus its future actions and perceptions. Applying this concept to robots results in a field of autonomous behavioural machines capable of operating in partially unknown and changing environments without human intervention. In the paper we describe the design and mechanical operation of the multi-robot system, the local positioning and communication system, ending with a discussion of the programmed behaviours and the final workshop outcomes. The results of this workshop will be used to inform the next stage of the technology demonstration in which the process will be scaled up.
Conference Paper
Full-text available
Designing for extraterrestrial life is a very up-to-date issue. However, there are many constraints in this kind of designs. Designs that provide the best solution can only be obtained by identifying these constraints very well. In this study, a design concept was developed for life in Mars by considering various constraints. This design consists of a kinetic system with pneumatic structure. The preliminary scheme of this structure, which was planned to produce as a prototype, was discussed in the scope of this study.
Article
Ionic liquids (ILs) have been considered as suitable media for the extraction of metals and other elements from the Martian regolith. The complex mineralogy of the Martian regolith and the innumerable possible combinations of cation/anion pairs in the available ILs pose challenges to the selection of most effective ILs to extract desired regolith elements. In this work, we utilized molecular dynamics simulations to fundamentally investigate the interactions between three major element oxides in the composition of the Martian regolith simulants, represented as alumina, hematite, and silica, and one basic and one acidic IL, i.e., 1-ethyl-3-methylimidazolium acetate ([emim][Ac]) and 1-ethyl-3-methylimidazolium hydrogen sulfate ([emim][HSO4]), respectively. We used potential of mean force (PMF) calculations to generate the free energy profiles of the cations and anions of both ILs on the oxide surfaces. Additionally, we generated their number density profiles on these surfaces to examine the structural features, including ion layering, at the IL-oxide interfaces. Our results indicate that both ILs have less favorable interactions with the silica surface and, hence, are less stable on it than on the metal oxide surfaces. Among the two ILs, [emim][Ac] shows more favorable energetics with the hematite surface than [emim][HSO4]. However, clear distinction cannot be made between the two ILs regarding their interactions with the alumina surface. Our calculated net interfacial energies of the IL-oxide systems further confirm our findings. The current work provides a general methodology to perform an initial screening of different ILs with respect to their potentials to selectively extract metals and other elements from the Martian regolith.
Conference Paper
Full-text available
http://www.hou.usra.edu/meetings/lpsc2014/pdf/1137.pdf
Article
Full-text available
Swarm robotics is an approach to collective robotics that takes inspiration from the self-organized behaviors of social animals. Through simple rules and local interactions, swarm robotics aims at designing robust, scalable, and flexible collective behaviors for the coordination of large numbers of robots. In this paper, we analyze the literature from the point of view of swarm engineering: we focus mainly on ideas and concepts that contribute to the advancement of swarm robotics as an engineering field and that could be relevant to tackle real-world applications. Swarm engineering is an emerging discipline that aims at defining systematic and well founded procedures for modeling, designing, realizing, verifying, validating, operating, and maintaining a swarm robotics system. We propose two taxonomies: in the first taxonomy, we classify works that deal with design and analysis methods; in the second taxonomy, we classify works according to the collective behavior studied. We conclude with a discussion of the current limits of swarm robotics as an engineering discipline and with suggestions for future research directions.
Chapter
An analysis of the history of technology shows that technological change is exponential, contrary to the common-sense “intuitive linear” view. So we won’t experience 100 years of progress in the 21st century — it will be more like 20,000 years of progress (at today’s rate). The “returns,” such as chip speed and cost-effectiveness, also increase exponentially. There’s even exponential growth in the rate of exponential growth. Within a few decades, machine intelligence will surpass human intelligence, leading to The Singularity — technological change so rapid and profound it represents a rupture in the fabric of human history. The implications include the merger of biological and nonbiological intelligence, immortal software-based humans, and ultra-high levels of intelligence that expand outward in the universe at the speed of light. You will get $40 trillion just by reading this essay and understanding what it says. For complete details, see below. (It’s true that authors will do just about anything to keep your attention, but I’m serious about this statement. Until I return to a further explanation, however, do read the first sentence of this paragraph carefully.) Now back to the future: it’s widely misunderstood. Our forebears expected the future to be pretty much like their present, which had been pretty much like their past. Although exponential trends did exist a thousand years ago, they were at that very early stage where an exponential trend is so flat that it looks like no trend at all. So their lack of expectations was largely fulfilled. Today, in accordance with the common wisdom, everyone expects continuous technological progress and the social repercussions that follow. But the future will be far more surprising than most observers realize: few have truly internalized the implications of the fact that the rate of change itself is accelerating.
Conference Paper
Two potential methods for lunar surface stabilization are: 1) sintering the regolith into a solid and 2) using heat or UV cured polymers. Sintering, a method in which powders are fused into a solid, has been proposed as a way of building lunar launch pads, roads and other building materials. Polymers are currently used by the military to stabilize sandy soils, and adaptations of this technology may be effective for lunar surface stabilization. This paper describes ongoing work at NASA Kennedy Space Center on these two technologies. A solar concentrator has been built to provide the heat source for sintering. Various solvent free polymers have been investigated. Results of physical testing, including load strength and abrasion resistance, on field and laboratory prepared samples are presented.
Article
The research goal of collective construction is to develop systems in which large numbers of autonomous robots build large-scale structures according to desired specifications. We present algorithms for TERMES, a multi-robot construction system inspired by the building activities of termites. The system takes as input a high-level representation of a desired structure, and provides rules for an arbitrary number of simple climbing robots to build that structure, using passive solid building blocks under conditions of gravity. These rules are decentralized, rely on local information and implicit coordination, and provably guarantee correct completion of the target structure. Robots build staircases of blocks (potentially removable as temporary scaffolds) that they can climb to build structures much larger than themselves.
Article
Complex systems are characterized by many independent components whose low-level actions produce collective high-level results. Predicting high-level results given low-level rules is a key open challenge; the inverse problem, finding low-level rules that give specific outcomes, is in general still less understood. We present a multi-agent construction system inspired by mound-building termites, solving such an inverse problem. A user specifies a desired structure, and the system automatically generates low-level rules for independent climbing robots that guarantee production of that structure. Robots use only local sensing and coordinate their activity via the shared environment. We demonstrate the approach via a physical realization with three autonomous climbing robots limited to onboard sensing. This work advances the aim of engineering complex systems that achieve specific human-designed goals.
Article
In current robotics research there is a vast body of work on algorithms and control methods for groups of decentralized cooperating robots, called a swarm or collective. These algorithms are generally meant to control collectives of hundreds or even thousands of robots; however, for reasons of cost, time, or complexity, they are generally validated in simulation only, or on a group of a few tens of robots. To address this issue, this paper presents Kilobot, a low-cost robot designed to make testing collective algorithms on hundreds or thousands of robots accessible to robotics researchers. To enable the possibility of large Kilobot collectives where the number of robots is an order of magnitude larger than the largest that exist today, each robot is made with only $14 worth of parts and takes 5 minutes to assemble. Furthermore, the robot design allows a single user to easily operate a large Kilobot collective, such as programming, powering on, and charging all robots, which would be difficult or impossible to do with many existing robotic systems.
Article
The unique properties of lunar regolith make for the extreme coupling of the soil to microwave radiation. Space weathering of lunar regolith has produced myriads of nanophase-sized Fe 0 grains set within silicate glass, especially on the surfaces of grains, but also within the abundant agglutinitic glass of the soil. It is possible to melt lunar soil i.e., 1,200– 1,500° C in minutes in a normal kitchen-type 2.45 GHz microwave, almost as fast as your tea-water is heated. No lunar simulants exist to study these microwave effects; in fact, previous studies of the effects of microwave radiation on lunar simulants, MLS-1 and JSC-1, have been misleading. Using real Apollo 17 soil has demonstrated the uniqueness of the interaction of microwave radiation with the soil. The applications that can be made of the microwave treatment of lunar soil for in situ resource utilization on the Moon are unlimited.