Early Days

 Unless some cataclysmic event occurs, we are not going to be building huge thousand-person colonies in space any time soon. It's just too expensive, with uncertain payoffs, unknown risks and very long time scales.

 The only way to get around that is to solve problems cheaply and use the available materials. For example, we have multiple rovers on the surface of Mars. These have operated for long periods of time and have taken photos and rock samples across an impressive track of travel in a very hostile environment. What could we do with a rover the size of the Mars Science Lab (Curiosity) on the Moon, with equipment for gathering resources?



 There has been a recent trend of small demonstration satellites (CubeSats and others) being launched as secondary payloads. Often a satellite does not need all the capability of the launch vehicle that delivers it. For a modest fee, other customers can add small payloads that tag along for the ride. Prices can be as low as $40,000 to launch a 1U CubeSat (that's 10 liters and up to 1.3 kg). Most commercial launch providers will handle add-on payloads like this for reasonable prices. Note that $40k is a lot more than the ~$3k per kg that sometimes gets quoted; that's because there are costs that do not depend on the mass of the object to be launched. For example, any payload has to be tested to ensure it will survive the heat, shock and vibration it will experience during launch; payloads also have to be proven safe for other payloads and for the rocket itself. These costs are trivial when spread across a five-ton communication satellite, but they are a major part of the price tag for small objects. Even so, the total costs are in the range of $100k and within the budgets of many universities and private companies.

 One concrete step we could take would be to collect many submissions for resource harvesting and related technologies. These would be packaged according to a standard (probably CubeSat-based and extended to 6U/12U and possibly 27U) and carried to the Moon in a carrier spacecraft. For a very reasonable price we could get some real, measurable results on various ideas for processing lunar dirt.
 Those ideas that have merit could be developed into several Curiosity-class harvesters (~900kg) and deployed along with a minimal base station. The rovers would explore the terrain nearby, harvest useful materials and bring them back for processing. The base station would provide electrical power, communications and processing equipment (plus night-side power). No provision would be made to go and collect them; if they produce enough useful material to justify collection then so be it.
 An equatorial landing site might concentrate oxygen, nickel-iron nodules and/or aluminum/titanium/magnesium. Polar sites would explore craters and potentially harvest water ice or other frozen volatiles in addition to oxygen and base metals. Depending on how advanced the concepts are, the site could potentially build its own tanks from basalt fiber or aluminum and fill them with oxygen. Proven results could inspire larger investments, while proven reserves ready to use on-site could inspire more ambitious manned missions.

 Another concrete step would be to take a similar competitive approach to greenhouse automation. Small deployable greenhouse units for testing automation as well as specific species performance could be housed in a carrier spacecraft and deployed in low or high Earth orbit, Lunar L1/L2 or on the Moon's surface. Those that do well could win predeployment contracts from NASA or other agencies for future missions. In places where carbon and hydrogen is readily available, automated greenhouse modules could be extended to producing methane, ethyl alcohol or even polyethylene plastics like Spectra from biomass. A truly automated production process that takes in light, carbon and hydrogen and spits out high-strength tether material is something to get excited about.

 The third area that is often mentioned but rarely included in serious plans is semiconductor manufacturing. It is often assumed that building solar panels on the Moon is a pretty easy task, yet we've not come close to performing such a feat. If the individual steps could be demonstrated in low-Earth orbit on small, low-risk demo satellites then we could more confidently proceed to a prototype factory on the Moon. This would also have applications for solar power satellites and captured asteroid exploitation.
 A solar panel factory on the Moon could put to use the materials harvested by a set of rovers, providing additional PV power for their base station in return. A system like this could start with very few trips to the Moon (potentially just one) and expand to large fields of solar panels to be used by other processes like oxygen extraction.
 In addition to solar panels, a number of other semiconductor devices are critical to expansion. LEDs are needed to add lit area to hydroponics. Solid-state rectifiers can be used to receive microwave beamed power. Components like power transistors, solid-state memory and microcontrollers would also be important for future expansion.

 Private companies would have incentive to participate as a demonstration of their technical capabilities to customers of space service firms and government space programs. NASA and other space agencies would benefit by maturing the TRL of a broad variety of technologies and materials for minimal cost. Competitive events like this tend to draw public attention, which is generally good for the space program as a whole and also good for science education.
 The development of a 'carrier' spacecraft would be useful for exploration at a number of destinations. It would be something simple and reliable that can be flown several times to save on development costs, yet can deliver a wide range of instruments and experiments that could be swapped out almost up to the day of launch. This standard frame that handles standard payload modules could be tasked to missions near Earth, Luna, Mars or further out in the asteroid belt with appropriate hardware added as payload modules, streamlining integration with instruments and components from other space agencies and private companies.
 Carrier demonstrations might include debris removal (carrying many expendable despin and deorbit sats), on-orbit assembly of large antennas and reflectors, delivery of many instruments to Mars/Phobos/Deimos or simply as a reliable frame for solar electric propulsion as a tug and later as an exploration vehicle.



 Once we move beyond these 'proof of concept' missions, we can begin the process in earnest. Early mining equipment will excavate, providing abundant geological and compositional data back to Earth, accumulating base materials and providing physical locations to place and bury habitats. Self-contained hydroponics pods will be deployed, providing raw material for the production of plastic as well as reserves of storable food (rice, beans, wheat berries; all dried and vacuum-packed). Manufacturing facilities will turn out solar panels, racks, cabling, O2 tanks, fuel tanks, flywheels (for energy storage), lunarcrete blocks and other useful things.
 NASA has been trying to put a nuclear reactor back into space since the 60's; powering an industrial site on the Moon would offer an ideal way to test a prototype with limited risk to humans while offering more benefit that simply demonstrating safe operation.
 As these sites accumulate resources, eventually it becomes profitable to build a low-capacity Lunar L1 or L2 elevator and use the tether material being produced on-site to add capacity. Regolith blocks, solar panels, energy storage devices and tanks would be hauled up the tether to add mass and functionality to the central station. At this point things have the potential to really take off and we can start envisioning serious trade from the Moon back to Earth.
 If at some point NASA or someone else manages to capture an asteroid into Lunar orbit then things become much simpler. Supplies of water and carbon greatly expand what is possible and make it easier to support a human presence. I don't mean a flags and footprints fiasco, but a permanent presence of mechanics, technicians and scientists.



 From here we can move on to large construction projects in Earth orbit, additional asteroid captures and eventually colonization of other bodies like Mars, Phobos/Deimos and Ceres. These activities would provide a steady stream of science data in addition to all manner of materials back to Earth. Earth satellites would be cheaper, which means voice and data services in many places would be cheaper, weather radar would be more accurate and Earth mapping would be more frequently updated. Large solar power satellites would finally be feasible, allowing nations to develop a non-carbon and non-nuclear baseload energy source. The number of telescopes scanning space would be vastly multiplied, to the point that we should be able to detect anything potentially approaching Earth. Sources of platinum and related metals would enable industry on Earth to expand the use of these materials as catalysts, reducing the costs of some chemical products (including gasoline) and potentially opening up more efficient means of generating hydrogen and purifying water.