1.0 - Introduction
This Wikibook serves several purposes. The first is a textbook-style introduction to production systems that feature high levels of self-expansion, integration, and automation. Their design builds on experience in existing fields of Engineering and Industrial Technology. A starter set specifically designed with these features we call a Seed Factory. A seed factory starts with a small set of flexible automated equipment, plus some amount of conventional shop tools. It can use local energy and raw materials to expand itself to a desired larger production capacity. The expanding factory also provides self-maintenance and useful products as outputs. Physically, the equipment may be at one or many locations, and under single or multiple owners. Regardless of their location and ownership, the equipment communicates and transfers inputs and outputs so as to function as a cohesive whole.
The combination of (1) a broad kind of expansion in size, complexity, range of outputs, and level of automation, (2) use of a starter set designed for self-expansion, and (3) using local inputs of materials and energy to supply the factory is a relatively new concept, so we adopt a new name for it. The particular name "Seed Factory" was chosen to show the relationship to both biology and manufacturing. Being new, no functioning seed factories have been built yet. The second purpose of this book is to support actual designs leading to working examples. We include several such design examples, and provide detailed notes supporting how they can be developed. We invite collaboration to work on these examples as open-source projects, with a goal of building individual machines and complete systems.
- Book Organization
The remainder of this section (1.0) starts by describing how self-expansion relates to previous ideas about replication, biology, and manufacturing. We then explore how the design of such a factory is different from a conventional one, and the potential advantages of this kind of production system. Finally we introduce some candidate projects to develop. Section 2 looks at the history and progress of related ideas, current work, and future technology needs and plans. Section 3 covers individual concepts and methods to better understand and design self-expanding systems. Section 4 then combines these pieces into a design process. Sections 5.0 through 8.0 take our candidate projects and develop them in more detail as design examples. The final section, 9.0, includes additional notes and reference material which are too detailed or too new to include in the main discussion, but which help support it.
Because Seed Factories are new, we don't have the level of experience to draw on that more established fields of engineering have. Therefore this book is necessarily incomplete, and will remain so until enough experience and working examples exist.
The ideas embodied in the seed factory concept have their own histories, and relationships to other ideas. It is worth identifying these key ideas and placing them in their larger context before describing them in more detail.
All factories can be self-expanding to the extent they make a product used in their own construction. For example, a sawmill produces lumber and the mill can be built mostly from wood. A conventional machine shop can make metal parts of the same kind used in its own machines. Conventional factories, however, do not include self-expansion as a primary goal. Rather, they are designed to produce a specific range of products efficiently and economically. Using these products internally is not something they were specifically designed for. When it happens, it is a side effect of factories needing to be made of certain materials. The factories which happen to make products with these materials can then use them in their own construction.
If desired, a machine shop can can purposely self-expand by making metal parts and then assembling them into more machines. However parts fabrication and assembly are only two of the steps in a longer production chain. Machine shops usually don't make the raw bar stock they use, or refine the ores to make the alloys for the bar stock. They typically don't produce their own power for the machines, either. A more fully self-expanding factory would do more of these steps from raw materials and energy supply to end products. We call one that can copy 100% of its own parts from raw materials Self-Replicating. This idea has been seriously explored since about 1950, starting with the brilliant theoretician John von Neumann.
Self-expansion is a broader concept than self-replication. Replication means making an exact copy of the same parts the factory is made of, and is often assumed to be a fully automated process. Self-expansion includes both less and more than just making exact copies. A seed factory starts with a smaller and simpler set of equipment which may not be able to copy itself directly. This is because a limited set of machines may not be able to make all the parts and materials from which they are themselves made. The starter set can expand to a physically larger and more diverse set of equipment by both making larger machines, and making new new machines not in the original set. This is in addition to making copies of some of the original parts. After expansion, it is able to produce a wider range of products, including even more diverse parts for even more new equipment types. After growing sufficiently, the expanded factory may now be able to produce much more than just copies of the starter set. This sort of growth and then making copies of a starter set is called "indirect replication" and is found throughout the biological world.
Operationally, a growing factory can make some of its own parts and materials internally. The remainder are supplied from outside the factory. As the factory grows, the additional machines and processes allow it to make a wider range of its own parts, and need fewer items from outside. It is not likely to reach 100% self-production. This is because some raw materials will be scarce in a given factory location, and some hard-to-make items are easier to buy from specialists than try to make yourself. Leading-edge computer chips are an example of the second category. It would be more economical to over-produce products that can be made internally, then trade them for the rare or hard to make items. The percentage of self-production then becomes a matter of economics and optimization.
A starter set is easier to design than a fully self-replicating factory. There are fewer machines since they use a smaller range of materials and make fewer kinds of parts. By choosing the most common materials to start with, like steel and glass, a few machines can capture a high percentage of self-production right away, and therefore lower the cost to get started.
Automation at a basic level is the use of control systems to operate equipment. A simple example is a thermostat (a temperature controlled switch) to turn a house heating system on and off. A variety of control systems have been used since the mid-1700's. In the 20th century, increasing use was made of electrical, electronic, and then programmable controls. The newer types were easier to modify and could carry out more complicated control tasks. More advanced automation allows equipment to mostly run itself, producing more outputs relative to people's labor. However, full automation is not yet practical, even for simple tasks. People still have to set the desired temperature on a thermostat. More interaction is needed for complex tasks like running and maintaining factory machines, even if they can run unattended for a time. Humans are very flexible general-purpose "machines". They can fill in for machinery and automation not yet added to a growing factory, or perform tasks not yet possible or economic to automate. Rather than assume a seed factory starts with full automation, we assume people are used where it makes sense, and it evolves towards higher levels of automation over time.
- Local Energy and Materials
As of 2013, 81.4% of World Energy Consumption was powered by fossil fuels, including most of the transportation of raw materials and finished products. We now know this use is unsustainable. We would prefer using local renewable energy sources and raw materials. Renewable energy directly substitutes for fossil fuels. Local material sources reduce the amount of transport required. Since most transport is currently fossil fuel-based, reducing it makes it easier to become sustainable, even before a shift to things like electric vehicles. For future seed factory applications in space, it takes a lot of energy to reach the lowest Earth orbits, and generally the farther you go, the more energy and fuel is required. Most satellites already produce their own energy in the form of solar panels, but larger and more distant space projects will benefit from using local materials too, as will avoid the large energy and fuel needs of shipping everything from Earth.
Biological seeds grow into larger plants using local materials and energy, and eventually produce copies of the original seeds. By analogy, a seed factory grows from a small starter set (the seed) to a larger production capacity, and can eventually make copies of the seed equipment. Both are examples of Indirect Replication, where the seed does not immediately copy itself, but first grows into a larger entity and then makes a copy of the original. Besides the analogy to biology, a seed factory is not restricted to mechanical production. It can use plant seeds to generate wood and fiber products, food, fuels, and chemical feed stocks.
A self-expanding factory may be thought of as a form of artificial life, because it exhibits many of the properties we associate with living things. These include homeostasis (maintenance of internal environment and state), internal organization, metabolism (converting materials and energy into parts for itself and wastes), growth, adaption (ability to change with time to respond to external conditions), response to stimuli (for example, production requests, or varying solar input), and reproduction (the ability to make copies of itself).
A factory is a purposely built place to do manufacturing processes, and a seed factory is therefore definitely a particular type of factory. We use the term "factory", rather than calling it a self-expanding machine, for several reasons. With the present state of technology a number of different materials and production processes are required, each of which is best carried out by a separate machine designed for it. So our factory will consist of multiple machines. Also, for the size and quantity of products we want to make, the mature set of equipment is closer to commercial factory building size than garage or desktop size. The starter set may be much smaller than this, but we think of it as the "seed from which a factory grows", in the same way an acorn is the seed from which an oak tree grows. A future generation design may completely fit within a shipping container or on a desktop. It might then be so integrated that Seed Machine would be a better description than a factory composed of multiple machines, but we don't think such an integrated design is possible yet.
In English we make a distinction between a factory, which tends to output many units of the same products, and workshops of various kinds, like woodworking and automobile repair shops. Workshops change what they are doing on a frequent basis, according to outside needs. Typically factories are also larger than workshops. Some of the self-expanding systems we describe in this book might better be described as workshops, because they grow to make a variety of items on demand and reach a smaller final size. For simplicity we will use the term "factory" to refer to all collections of production equipment, rather than use two names. We recognize, however, that mass production is different than per-job work.
- Economic Development
In addition to the words that make up the label "Seed Factory", self-expanding systems are related to economic development. Seed factories can be useful in less developed places on Earth, or regions in space which start out completely undeveloped. By providing large amounts of physical products and energy from a small starter set, they can help build up an area efficiently. Mere physical outputs alone, however, are not a complete solution to developing an area. That requires integrating social developments like health care, education, and legal and civil rights. Modern design processes incorporate these other factors by including them as design requirements, functions, and flows. For example, waste outputs and safety hazards can impact the health of a community. Wastes and hazards can be limited by imposing suitable requirements on the factory in the design stage.
Saying we want a factory that is self-expanding does not tell us what features should be included, or how it is different from a conventional factory. Collectively we call these features and differences the Design Approach. We list some of the elements of the design approach here, and will go into more detail later in the book:
- Growth, Adaptability, and Integration
Traditional factories tend to be fixed in capacity for extended times, and make a limited range of products. Therefore a single design optimized for those conditions is a reasonable approach. In contrast, a seed factory can grow continuously from the starter set to an expanded capacity. As it adds a wider variety of machines and processes, it can both make a higher percentage of parts for itself, and a constantly growing range of products. Therefore the evolution of the factory over time must be planned for in the original design. The future, however, cannot be entirely known in advance. End users may want different products, or a new production process might be developed. To adapt to these uncertainties, features like flexibility and modular design become more important relative to production volume. No matter what size it is, a seed factory should be designed as an integrated whole, rather than as a collection of machines that just happen to be housed in the same factory building. This means considering input and output flows between different production processes, using wastes from one process as inputs to another, and recycling items where possible. This is in addition to conventional design that tries to optimize the individual machines that perform each step.
- Automation of Change
Automation is widely used in modern manufacturing. However, reaching a high level of automation with a constantly changing factory is a new challenge. Factory planning has traditionally been a human design task. With a growing factory such as this, we would also like to automate the planning process. One way to approach this is to describe all the factory equipment and end products in a library of design files. The files not only specify the shapes of parts to make, but also their input material and energy needs, and the various processing and assembly steps. A request to produce a given item then gets converted to a series of tasks based on this data. The tasks are assigned to either automated equipment, humans where the task is not automated, or to purchase items that cannot be made internally. Additional production orders are assigned backwards from final assembly to earlier steps in the production chain. These make more inventory or other supplies for the later steps, all the way back to raw materials and energy supplies. Expanding the factory by adding a new machine is then treated the same as any other product, as a series of steps that lead to final assembly.
The production software would recognize what equipment is available in the factory at any given time. So, as the factory matures and develops, a given product design would generate a different production flow, based on the current capabilities. If there is not enough total capacity in the factory for all the production orders in progress, the software might also generate a request for more equipment to expand the factory. This would get inserted along with all the other production work. So the factory would adapt to current production needs. The design of all possible future equipment and products is not necessary or desirable at the start. You can build a seed factory with a limited set of design files covering the starter set of equipment and early end products. As designs for new products or factory equipment are developed, they can be added to the design library, and produced when needed.
We think that modern production designs need to go beyond just optimizing production volume and cost, and account for additional factors such as sustainability. Living plants mostly grow from commonly found energy and materials in the local environment, and biological systems have a high degree of materials recycling. We would like to copy these features, so that our factories can have the kind of ubiquity and sustainable time scales found in the biological world. The factories must be consciously designed for this - many existing factories use scarce resources with little, if any, recycling. Ways to design for this include making it an explicit design requirement, and considering the relationships between processes. For example, making cement gives off Carbon Dioxide, and plants in a greenhouse might want to consume extra Carbon Dioxide. It makes sense to use a waste product from one process as an input to another, but looking for ways to do so must be a regular part of the thought process during design.
- Complex Systems
Seed factories rely on advanced industrial automation, integrated flows, and continuous growth and change via design and self-expansion. The different processes and manufacturing steps combine elements from mechanical, electrical, chemical, biological, and other engineering fields. There are also multiple design goals we want to meet at the same time. This makes them quite complex projects to design and build. The Systems Engineering field has developed since the mid-20th century to manage such complex projects. We therefore adopt systems engineering methods as part of our design process. We also borrow design tools and methods from other fields, like industrial technology and building construction where appropriate. To help manage complexity, we also don't try to design everything at once. Instead, we follow an incremental approach, starting with smaller and simpler systems, and adding to them a little at a time.
Aside from purely intellectual interest, a practical question is why should anyone build a self-expanding design, rather than a conventional factory? The answer includes the technical, economic, and personal advantages a seed factory approach brings. If these advantages are large enough, then people will use that approach as a widespread production method. We have identified a number of features of seed factories that we expect will be advantages, and list them below. More work is needed to prove they actually are advantages, and to quantify by how much. We hope to make progress on this question in the course of writing this book, and by designing and testing equipment for such factories in the associated Seed Factory Project.
- Integrated Automation
A self-expanding factory should be more productive and less expensive than conventional factories. The design would include an increasing variety of computer/automated/robotic elements. This reduces how much human labor is needed for a given output, lowering that element of cost. Multiple production steps from raw materials to end products are brought together in one location. This allows automating the transfer between production steps, as well as the production itself. Compared to conventional specialized factories, it eliminates the packing and shipping between them, and the energy and labor needed for transport. The automated transfer and reduced transport distances should also significantly reduce cost. Self-production lowers the initial set-up cost for the factory, since you only need to purchase a subset of the equipment. New equipment can be built as time is available between products for sale. This helps keep factory utilization high, and minimizes idle capital and overhead costs.
The level of automation scales with the growth of the factory. In the early stages, a necessary machine or process might not be available. In that case the process flow would call for ordering a part from outside, or ask a human to do a manual task. When the factory is more mature, the same task might be entirely automated. Full automation is probably not possible at present, but we think that computers, robotics, software, and sensors are advanced enough that a highly automated and integrated factory can be built. We also think existing technology enables running such a factory mostly on locally available raw materials and renewable energy. This should make it both sustainable and low cost to operate. It therefore should compete favorably with traditional factories.
A seed factory, within the capabilities of current technology, can grow to nearly make full copies of the starter equipment. So after the first one is built, you can get a nearly exponential growth in capacity. It is flexible and general purpose by nature, able to make any product it is fed instructions for and within its capabilities. This includes a variety of useful end products, new equipment for diversification, and more seed factories if desired. Although organized into different processes and machines, the factory as a whole is designed and operated as an integrated system with unified control. Each part of the factory contributes to the maintenance and operation of the other parts, as well as to useful final products. So it is substantially self-supporting and independent. This kind of growth is especially useful where local production is lacking or non-existent.
A seed factory is more portable than a conventional factory. The most compact version is just the set of design files for the starter set, expansion equipment, and end products. The design files would include instructions on how to build the starter set from easily available sources and equipment, and then how to progressively expand it. In this form it can be distributed anywhere at low cost. A more complete and ready-to-use version would include some hard to make parts and materials along with the design files. Common local supplies would be added to these to build the complete seed factory. An even more complete version would arrive in shipping containers, immediately ready to work. These portable forms contrast with the traditional "site-built" factory, which is fairly immobile. In outer space, portability is especially useful, because of the high cost of transportation.
A seed factory does not have to exist in a single physical site, like most traditional factories. If desired, it can operate as a distributed production system connected by computer networks. However, automated transfer between different machines, or delivery of parts and supplies between them, is made easier if they are physically close together. With modern communications the owners and operators can also be distributed, controlling some operations remotely, while the hardware is in close proximity for efficiency. The flexible layout is an option enabled by modern technology that was not available in past decades. It lets the owners arrange things how they want, rather than being forced to put the people and machines in one place because it was the only way possible.
The economics of a seed factory should be better than a conventional factory. At first it cannot make 100% of the parts and materials it needs. It produces a surplus of the things it can make, and sells them to pay for the items it cannot supply internally. As the factory grows towards maturity, with increased process diversity and capacity, it will make more of it's own items and need less from outside, thus lowering the cost of production. To the extent the factory can build itself, rather than buying all the equipment directly, it requires less capital than directly building a conventional factory.
Traditional factories reach low unit cost by economies of scale and mass production. A mature factory grown from a seed reaches low unit cost by reduced capital, automation, integration, and using inputs closer to raw materials. Conventional factories tend to be large and unable to operate until completed. Therefore they need a lot of capital investment which must wait until production starts for a return. The capital tends to come from large and patient sources, who are usually not the same people who eventually work in the factory. Growth from a starter set makes it easier for the workers to also be the owners. As owners they will be more secure in their jobs, and more willing to invest in themselves and the factory for the long term.
In addition to proving design feasibility and technical advantages, the seed factory concept should have suitable practical applications. Different applications will likely need different starter sets and different growth paths. So identifying and designing for these applications will help discover what parts of the factory are universal and what parts are particular to a given use. Functioning seed factories have not been built yet, so we can't point to existing applications like in more experienced engineering fields. Instead we will present several future examples in the later sections of this book. They will serve to show the range of possible uses, and differences between designs prepared for different purposes. The process of developing each example is also a guide for how to design for other applications not covered in the book.
Our current examples are drawn from two sources. One is a proposed program to upgrade civilization and expand into space using seed factories as a core method ( To Mars and Beyond, Eder, 2015). The other is a list of sample industries selected for concept exploration studies. The study process is to work backward from those industries to identify starter sets that will lead to their needed equipment and products. These examples are developed in more detail in later sections of the book:
- Personal Production
Our first example has the goal of making a wide range of products for a local community of owner/operators, such as food, building materials, and utilities. It begins with a network of people interested in working together, conventional workshop tools, and a starter set of seed machines. At first, production is at hobby and home-use scale, to keep initial costs within reach. Later equipment is scaled in size and diversity to reach larger capacities. This example is personal in the sense that the equipment owners can make things for their own use. It is not like a desktop personal computer as far as all the components having a single owner. The collection of equipment eventually becomes too large and expensive for one person to have at home, and too complex for one person to operate. The solution is larger and more dedicated production places, and multiple owners with different skills who work together to make what is needed.
There are two approaches for the owners to work together: distributed and centralized. In the distributed approach, each person owns one or a few pieces of equipment. A production request is then distributed as needed to the people and their equipment, who work together to make the product. In this approach, ways are needed to settle up resources and costs used by people for each others needs. In the centralized approach, the equipment is in one or a few larger places, with shared ownership. A mixed approach is also possible, part distributed and part centralized. As production capacity grows beyond personal needs, it can start to operate like a commercial workshop, building items for outside customers in addition to the owners. At the hobby and home-use scale, participation is mostly part-time, and the network of owners' main support is from conventional work. As production grows to commercial levels, some of them can transition to full-time self-employment and have the majority of their physical needs supplied from within the network.
- The "Makernet"
A Location in our concept is an area small enough that people can easily travel to work in person, and exchange products with each other. An example would be a US metropolitan area. In the Personal Production example, one location can then contain multiple home and commercial sites where work is done. This example extends the distributed approach to multiple locations all over the world. Production sites become network nodes that use the Internet to communicate with each other. We call it a "Makernet" because it makes things across the Internet. Automated production machines can communicate with each other directly, with users/customers, and with remote operators. Since modern communications are worldwide, the network does not have to be restricted to a single location. Shipping physical items long distances involves significant cost, energy, and time. Therefore Makernet production will tend to be concentrated on higher value/mass ratio items. Electronic designs for new machines can be sent anywhere at low cost, while crushed stone for concrete, which is heavy and cheap, will tend to be limited to a single location.
Operationally, a customer may submit a product order through a website. Production requests are then routed to various network nodes, who supply the materials, fabricate parts, etc by the most efficient path. Eventually the finished item is delivered. Production nodes can contain any number of machines and processes. Since the machines can be made by already existing nodes, a single seed factory starter set can spread everywhere over time. For node operators, the Internet allows people to work remotely from the actual hardware. This has potential advantages in commuting cost and flexibility in hours and location. In the previous example of a local community at a single location, much of the interaction can be in person or by voice. The Makernet approach emphasizes software and networking, and starter sets would be optimized for low mass, so they can be shipped farther at low cost.
- Industrial Production
In this example, the end goal is large scale production of more specialized products, much like conventional factories. This example begins with either a smaller-scale commercial network as in the first example, or starter sets specifically designed to grow to industrial scale. In either case, growth by self-expansion is a less expensive way to reach the desired production capacity. In one version, the starter sets are mobile, and are used like construction equipment. They are brought to a new industrial site and start making parts and equipment. Once the new factory reaches sufficient capacity, the seed factory moves on to the next project. In another version, the commercial-scale network grows in place to the larger industrial scale. Individual industrial sites grow towards scale and efficiency rather than diversity of outputs. Partly this is because it becomes harder to fit all the equipment and operations in one place. Individual machines can optimized for their specific tasks. They are then easier to automate because they do fewer different things each. Different sites can still operate as a network, and make things for each other as needed. The larger scale output goes well beyond the original owner's needs. They therefore can sell products directly to a wider market, or sell production shares to people who otherwise don't participate in site operations. Production shares entitle people to a share of the products on demand, either without further payment, or just to cover raw materials. Industrial-scale sites can also use conventional factory financing via capital markets if internal funding is not sufficient.
- Remote and Difficult Conditions
The goal for this book's last example is production in remote locations, or where environment conditions are difficult or extreme. In such locations, the benefit of self-expanding systems is not having to bring all the infrastructure with you. Instead, you bring a starter set and produce most of the rest locally. Examples of remote and difficult locations on Earth include deserts, polar regions, and the ocean surface. More extreme locations include high altitudes, ice caps, deep underground, or underwater. Current civilization mostly occupies a thin layer on about 13% of the Earth's surface. Extending it horizontally to the remaining area, and vertically beyond the first 100 meters would vastly increase available resources. The key to this is affordability, which self-expanding systems can help with.
The more difficult and remote the location, the higher the cost of transporting everything there, and the more incentive you have to use a small starter set. High levels of automation and productivity would make it possible to live and work in these difficult locations by providing the infrastructure and resources needed. Remote controlled operation allows building up capacity without necessarily living there. For example, the Sahara Desert may be a great place to make solar panels, because of abundant sunlight and sand as inputs. But people may not want to live there full time, and working remotely can allow that choice. Conversely, if technology enables living in a remote location, it might be desirable for some. Since other people can't live there, land would be cheap, or it might scenic or have other features people would prefer.
The most remote and difficult locations are in space beyond the Earth. Synchronous orbits, where many satellites operate, are twice as far as any two points on Earth can be, and the farthest known Solar System orbits extend ten million times farther than that. The space environment has a number of difficult or hostile conditions for people and life as we know it, such as lack of breathable atmosphere and water, high radiation levels, and low gravity. On the other hand space, contains vast amounts of untapped energy and material resources. If seed factories and self-expanding production can help make these resources accessible, the potential gain for civilization is enormous. Since the conditions in space are so different, and involve different engineering specialties, we reserve discussion of this type of location to another Wikibook, Space Transport and Engineering Methods.
This wikibook is being developed as part of the Seed Factory Project, which is an open-source collaboration to develop the technology and hardware for seed factories. The original author is Dani Eder, 6485 Rivertown Rd., Fairburn GA, 30213, user Danielravennest on Wikibooks, and email firstname.lastname@example.org. Other contributors are welcome and can choose to add their names and contact info here if they wish. Otherwise the history tab on any page indicates the source of editorial changes. If you contribute to the book, we ask that you provide sources for your data and calculations, so that others can check the work.