This page considers a possible approach to developing a low-mass scientific payload for extra-solar use. The ideas expressed here are far from original and a few relevant links have been included to items thrown up by a cursory web search but if you know of others, please met me know.
A pre-requisite for any such scientific investigative package (a 'probe') is that it arrive at its destination in a functional condition. Using only passive solar-sail technology, speeds in excess of 0.1% of the speed of light can be achieved but will still require over 3,000 years to reach even the closest star, and for those systems currently known to include planets, journey times in excess of 100,000 years may be required. Current technology cannot guarantee that lifetime for any single item, nor is it likely that this will be approached in the foreseeable future.
There is however, no need to try to build a single piece of high-reliability equipment. Such a target can easily be met by a system which is designed to include a high level of redundancy. Of course over time attrition would still reduce any system to a non-functioning state unless failed parts can be restored to operation. This requires that the design include the ability to repair or replace any part of the system and no part or sub-system can be considered exempt from this requirement. The ultimate limit on the life of a system composed of many parts comes from the finite probability that a random fault will occur in every part simultaneously. To ensure that the system always achieves close to that theoretical limit, causes of failure other than random events must be eliminated. Wear occurs in all mechanical systems and in semiconductors migration of material and solution of surface features into the bulk material place a finite life on any structure. A robust design will therefore include a 'planned maintenance' schedule which breaks down each part and replaces it with a new version well before the end-of-life point is reached.
An obvious problem is that, here on Earth, we are accustomed to production of semiconductors and other components requiring large manufacturing facilities. These are obviously inappropriate for a low-mass probe. In addition, the power available in deep space is severely limited and can be considered to be that which can be provided solely by on-board systems using fuel carried from launch. It may be possible in the future to develop a collection system for interstellar hydrogen and a pocket-portable fusion power plant, but these are not currently available and do not seem to be imminent. Construction techniques must therefore be able to construct any required part using power levels in the microwatt range.
There appears at this time to be only one technique which is becoming feasible in laboratory conditions that holds the prospect of meeting these requirements - that of atomic level construction or nano-technology manufacturing. By adapting the Scanning Tunneling Microscope (STM) it has been demonstrated that individuals atoms can be manipulated with a precision of a few nm and development of this should bring sub-nm accuracy within a few decades. Of course, the same problem arises if we think of using a conventional STM but the key feature is the use of a very small probe tip to get the required resolution. This has been done so far by attaching a carbon nanotube to the tip of a conventional STM. The same result could be obtained by designing a robot having a manipulating arm which is itself of size nanotube. The voltages needed to effect movement or removal of atoms from a surface are in the range of tens of millivolts to a few volts, well within the capabilities of conventional electronics, although at such small dimensions, field strengths will be higher and careful design will be needed.
There is therefore the prospect of building items one atom at a time from stocks of raw material. This would be a slow process and the time taken could only be reasonable if either the item to be made was very small or a large number of robots cooperated on the construction. The primary requirement is that probe be self repairing. A single robot should then have the capability to manufacture a duplicate of itself given sources of power and raw materials. The design of the robot should then aim to meet this in as small a structure as possible to minimise the construction time. Once a robot builds a duplicate of itself, the process can repeat and the total number of robots will rise exponentially with twenty construction cycles would provide over a million copies of the original.
Previous work has shown that data sizes of the order of 500,000 bits may suffice to define a robot capable of self-reproduction. Excluding the possibility of using data compression techniques to reduce this value, and assuming 10 atoms per stored bit, the minimum size just to hold the template is then of the order of 5 million atoms. Allowing a factor of 4 for the mechanical aspects of the robot, the mass of an individual would then be of the order of XXX micrograms.
Once an adequate number of robots has been produced, other larger structures can be manufactured in a reasonable time by designing the robots to act cooperatively above some threshold quantity. Coordinated group behaviour arising out of characteristics of an individual is already being investigated in the "Artificial Life" approach to programming. This is another area where development is required but the basis of the technology has been demonstrated.
The key difference between self-duplication by an individual and construction of larger and more complex items by an ensemble of robots is the amount of data required to define the product. Self-duplication could be based solely on information held by the robot, its own structure serving as the template. For other products, either an external source of information is required or the size of the robot must be increased dramatically to hold the instructions for all possible products, an approach which goes against the original target of keeping the robot size as small as possible. This also means that the range of possible products if the size .
Once two-way communications has been established, even with a response time of several years, there is no limit to what can be built remotely, given sufficient intelligence in the autonomous construction systems.
Back to the main summary.
A Java Solar Sail Calculator.