MM incentives
From Wise Nano
Contents |
Incentives and Applications
It is hard to imagine the extent of what could be done with advanced nanomanufacturing. Clearly, products could gain radically in performance and efficiency. This section argues that in addition to higher performance, the cost and time to develop products could drop significantly, and manufacturing costs could drop precipitously. The final subsection describes a few applications of the technology that are likely to be influential.
Rapid R&D and deployment
Manufacture of prototypes is currently a costly process. Rapid prototyping machines can help, but so far they can only make passive components, not integrated products, and the machines themselves are costly. Assembly is still required. Manufacture of a prototype product also takes substantial time.
Today, high volume manufacturing may require overhead including expensive molds, training of workers, and procurement of supplies. The design of products intended for high-volume manufacturing must take this into account; the product must not only be useful, but also manufacturable. Designing the manufacturing process is a significant part of the total design cost, in addition to the constraints it places on product design.
A nanofactory would be equally well suited to rapid prototyping and high volume production. A prototype could be produced in an hour for a few dollars per kilogram. This would allow rapid testing of new product ideas and designs, more like compiling software than like today's prototyping process.
Once a design was approved, it could immediately be put into production. Depending on how the nanofactories were deployed, production could be at point of sale or even point of use. Warehousing and shipping would not be required, substantially reducing the expense and delay of product deployment. Less reliance on economies of scale would allow efficient test marketing, potentially reducing initial marketing costs. Lower costs for R&D, and far lower costs for initial deployment, would allow greater tolerance of failure and thus more innovation.
Nanoscale and microscale designs could be developed by a rapid genetic algorithm-type process with physical evaluation of the fitness function. An array of varied designs could be built and tested, and the results used to refine the design parameters and specify the next test array. Rapid construction and testing of millions of components would allow physical implementation of genetic algorithm design methods. (Note that the products would not build copies of themselves—there would be no self-replication. The nanofactory would build each successive generation of products.)
Low cost of manufacture
As a general rule, prices are driven by a balance between demand and scarcity. A nanofactory capable of building a duplicate in an hour would be able to support a rapid increase in the number of nanofactories to any desired level. There would be no scarcity of manufacturing capacity; if nanofactories ever became more valuable per kilogram than their cheapest products, they would be used to manufacture more nanofactories. Even if the feedstock is relatively expensive, this argument implies that products with high value per gram such as computers and pharmaceuticals would not lack for manufacturing capacity. Of course, this argument ignores sources of artificial scarcity such as patents.
Although early nanofactories might require expensive feedstock and consume large amounts of power, a combination of deliberate design processes and genetic algorithm approaches could produce rapid improvements in nanofactory component design, allowing the use of simpler feedstock. Similar rapid design effort might be implemented to develop nanofactory-built chemical processing plants, reducing the cost of feedstock. Because small designs could be built in large arrays, the processing system could use any convenient combination of mechanosynthesis, microfluidics, and industrial chemistry.
If production capacity became non-scarce (at least to the patent holders) then it is not obvious what resource would be scarce. Lightweight solar collectors could produce a plentiful supply of energy. If the products include solar collectors, then energy would not be limited. (The active component of a solar collector can be quite thin, and the structure can be low-mass; a structure massing a kilogram per square meter would recover the energy cost of its manufacture in a day or so. Thermionic solar cells have been built out of CVD diamond.) Feedstock would be small carbon-rich molecules; carbon is readily available anywhere on earth, and nanofactory-built equipment might be used to process cheap carbon sources into feedstock. Nanofactories would not require special skill to operate, and would not require working conditions that would be expensive to maintain.
With near-zero labor costs, low environment and capital costs, and moderate energy and feedstock costs, there is no apparent reason why the per-item cost of production should be more than a few dollars per kilogram regardless of device complexity. Of course, the cost of design (including the amortized design of the initial nanofactory) will be far from trivial. This contrasts sharply with the value of the products to consumers—by today's standards, nano-manufactured computer components would be worth billions of dollars per gram. There would be a huge incentive for profit-taking; it is not at all obvious how soon consumers would see the full benefits of the new manufacturing technology, even if its military implications did not cause it to be restricted.
Applications
A general-purpose manufacturing system capable of making complete high-performance products will have many applications. These include massively parallel sensors and computers, military force multiplication, wholesale infrastructure replacement, ecological footprint reduction, and aerospace.
The ability to build kilograms of fully engineered nanoscale products means that vast numbers of sensors could be produced at near-zero cost. These could be integrated into one structure for optical or medical data-gathering, or could be incorporated in small distributed sensor platforms. More compact functionality, easier fabrication, and more efficient use of power would give nano-built sensor platforms a significant advantage over MEMS technologies.
The ability to gather large amounts of physiological data (e.g. chemical concentrations or electrical potentials) in real time using a sensor array small enough to be inserted into the body without damage would be a huge help to medical research. Early and accurate detection of health conditions would help in the mitigation of many diseases. Early detection of adverse reactions to treatments would allow doctors to design more aggressive and experimental treatments without compromising safety; it might even be possible to bypass clinical trials entirely. At the other extreme, better understanding of causes, effects, and feedback loops in the body would allow more subtle and less invasive treatment. Cell-sized machinery raises the possibility of cell-by-cell interventions, even genetic interventions, with more control and flexibility than current therapies. Interfacing to neurons, for both sensing and controlling and controlling neural signals, could be done more delicately and on a far larger scale than with today's electrode technology.
Massively parallel computers would be one of the first products of a nanofactory development program. This would have applications in simulation, sensor array data processing, data-mining, pattern recognition, symbol manipulation, and neural networks. Precise construction could also be expected to be useful in building quantum computers.
Portable high-volume general-purpose manufacturing of advanced products would greatly increase the flexibility and power of military activities. Sensor and computer improvements would greatly improve situational awareness. Manufacture of products at the time and place they were required would improve logistics and transportation. The ability to build high-performance computers and actuators into any (or every) cubic millimeter of products would allow new kinds of weapons to be developed. The ability to inexpensively and rapidly build, test, and deploy new weapons would accelerate progress in weapons technology.
Low cost, high throughput manufacturing could be used to build very large quantities of product. Inefficient products, and even networks of inefficient products, could be replaced with little manufacturing effort. Of course replacement depends on many other factors including design effort, installation effort, political resistance, compatibility, and consumer acceptance. But infrastructure is largely invisible to consumers, is often administered centrally, and is often aging or inadequate. Infrastructure may provide many suitable targets for widespread deployment of nanofactory-built products.
As the efficiency of infrastructure is improved, and the ability to monitor the environment increases, it will be increasingly possible to reduce humanity's ecological footprint. Accidental or unnoticed damage could be reduced, and the consequences of deliberate activities could be better known. Mineral extraction and processing activities, including fossil fuel consumption, could be reduced. Water could be used and re-used more efficiently. See Fig. 7. Even something as simple as widespread use of inexpensive greenhouses could save substantial amounts of water, topsoil, pesticides, fertilizer, and land area.
Fig. 7. A water filter with .3 nanometer pores (on left) would clean water down to the atomic level with minimal pressure drop due to drag. Such thin membranes would need to be supported by struts (center). Larger pore sizes are possible (on right). The very smooth surface on top would reduce fouling.
Aerospace hardware depends on high performance and light weight; this is especially true for spacecraft. Orders of magnitude improvement in material strength, power density of actuators, and computer performance, make it reasonable to think of designing rockets with hardware mass a small fraction of payload mass. Lightweight inexpensive hardware makes it easier to design combination airplane/rocket systems. It might even be worthwhile to include an efficient gas processing system and fill a collapsible liquid oxygen tank after launch. Combination airplane/rocket systems capable of reaching orbit could be much smaller than rocket-only systems. Today, orbital access is expensive due to minimum rocket size, high construction cost, and the additional work required to avoid expensive failures. If spacecraft were smaller and failures were substantially less expensive, then R&D could proceed with less deliberation and more testing of advanced chemical and non-chemical designs.
