Ideas and Notes

Dani Eder

The Seed Factory Project,

6485 Rivertown Rd, Fairburn, GA 30213

danielravennest@gmail.com

16 August 2021

Note: This page is a notepad for recording new ideas developed by the author, but which need more work for a paper or to include in a technical report. It is sorted by topic. As ideas get incorporated elsewhere they will be deleted from here.

1.0 -Self-Improving Systems Generally

(includes Starter Sets, Bootstrapping, & Self Expansion)

Alternate titles for self-improvement: methods for..., principles of..., applications and projects for...

1.0.1 - Systems in General

Systems in general can be measured along two axes (see figure). The first axis measures how good the system is according to parameters chosen by people. Examples for a factory might be economic value of production output or factory floor area, where higher values are considered better. Systems can either be improving, stable, or declining on this axis, and a rate of change can be determined at various points in time. Thus a system over its life may start out improving, reach a stable state, then decline. The second axis is to what degree system changes are from outside vs internal causes. A grain silo changes only due to outside causes - people putting grain in or taking it out. On its own, the silo does not change how much is being stored. A group of people who are starting a new company are largely self-acting. They can decide on their own to add or remove people from the group, assign tasks to be done, what products to make, and how to make them.

Our interest is in "Self-Improving Systems", which are high on both axes. They are designed to get better over time, and are largely driven by internal actions. In the production arena these systems can include both people and equipment, but self-improving systems can occur anywhere we can measure their goodness and degree of self-action. For example, a solar system outside our own can evolve over time as the central star and family of planets change on their own. We humans can apply "fitness for human habitation" as a parameter to measure how good this solar system is, and this value can change with time.

Factories That Can Improve Themselves

A typical factory, like one for making automobiles, makes a fixed range of products, at a more or less fixed rate, and is not designed to improve itself. This section covers the general idea of production systems which can improve through their own action, at whatever scale, purpose, and level of technology. They do this by making items to use internally, in addition to making items for others like any other factory. This results in self-improvement, expansion, growth, and upgrades of their own equipment. In turn, the new capacity allows more production for others, and greater capacity for further improvement. This can result in a super-exponential rate of change - not only increasing by a fixed percentage per time (exponential), but increasing the rate of change over time (super-exponential). The accelerating change comes from incorporating better tools and equipment that are more efficient and faster.

Accelerating self-improvement has great potential for the future of civilization. This provides motivation to study and pursue this kind of system. They have large-scale economic implications, and will therefore involve politics and advocacy to properly implement. These aspects are covered in subsections below, followed by accumulated descriptions, ideas, and names for these systems. After covering the general topic, later sections cover specific scales and applications of this idea.

Factories have always been able to change by outside action. They go from not existing at all to built and functioning by way of outside resources such as money, land, construction materials, labor, and machinery built elsewhere and delivered. They have also been able to change through trade. Sales of products they currently make are used to fund improvements. Some are able to partly improve themselves, when the product they make happens to be useful for that task. A steel fabricator who makes structural parts for buildings can use those parts to add to their own buildings. Companies who make machine tools or robots can use those products in their own factories.

The kind of production systems we consider here are intentionally designed for self-improvement, generally with an increasing ability to do so over time. A starter set of tools and machines may not have any ability to make items for it's own use. For example, a set of garden tools can help produce food in a backyard garden, but may not be able to make any better garden tools. A surplus of food can be traded for some carpentry tools. Those can be used to build raised beds, a greenhouse, or a potting shed, improving the garden's capacity. A larger set of tools in a home workshop or makerspace may be able to make some items for itself, but not copy all the tools. After a period of growth and upgrade it may become able to not only copy the original set, but produce new starter sets of different and better types, which can then follow their own growth paths. This sort of evolution towards greater self-change can be planned for ahead of time.

The garden tools, larger workshop, or community makerspace can't produce change on their own. If you include people as part of the system, change becomes possible. People are first able to change themselves, such as learning new skills or building muscles. They can then use the tools, workshop, or makerspace to produce changes and improvements to them. A collection of machines that use automation, robotics, software, and artificial intelligence (i.e. smart tools) can potentially change themselves without people, or with fewer people than "dumb tools".

Seed Factories and Self-Improving Systems

A factory system that can help improve itself from a starter set we call "Self-Improving". The starter set is called a "Seed Factory", in the sense that it is the seed from which the rest of the factory grows. The ability to self-produce its own parts can range from 0 to more than 100%, and can change over time. Seed factories are generally much smaller and simpler than the final factory. At first, they require inputs of ready-made parts and materials, and can only produce a small range of finished items. As these accumulate, they can make more items internally and need less from outside, relying on internal sources of energy and raw materials.

"Self-Improving" does not mean fully automated. People, with their skills and knowledge, are a part of self-production. Smart technologies like automation, robotics, software, and artificial intelligence can partly replace the need for people, but we are not at the point where such systems can run entirely by themselves. "Factory" does not mean a single central building where all the work is done. With modern communications and transportation, the work can be distributed across multiple places. So long as the work is coordinated and the desired products are made, it can still be considered a factory-type production system in the sense of making large amounts of finished products.

We can summarize the Seed Factory idea in the following way:

Seed Factories are systems that can bootstrap from a starter set of people and machines to whatever size you need. They can grow with the help of outside sources, but are designed to use their own tools, materials, energy, and knowledge where possible. This can include adding smart tools, local energy and raw material production, and applying process and design knowledge to improve the system. A mature factory grown from a seed can reproduce by making new starter sets.

[TO BE MERGED]

A seed factory is a starter set of tools and machines. They are used to make more equipment for itself, until they can make new starter sets. This is like how living plants grow from seeds and eventually make more seeds. Like any other factory, they also make useful products that people need and want.

A Seed Factory Network (MakerNet) is a distributed production system. It bootstraps from a group of people, with starter tools, resources, energy, and knowledge. They use their skills and equipment to expand and upgrade the network. This includes adding "smart tools", which use automation, robotics, software, and AI. Then the network can mostly run itself, making everything they need and want.

"Smart Tools", which use technology like automation, robotics, software, and AI, are making self-reproducing factories possible. These will enable cheap development of both the rest of the Earth, and space. That's because products made by self-reproducing factories that can mostly run themselves will approach the cost of raw materials, and in undeveloped areas raw materials are cheap.

1.1 - Technology Motivations

Reasons to study and pursue this kind of technology include

Enabling a post-scarcity civilization using self-expanding automation, renewable energy, and abundant raw materials
Upgrading civilization to where everyone has a high quality of life
Creating an open-ended future, rather than one constrained by a finite subset of the resources on Earth
Bootstrapping a better society, including fewer parasites who feed off scarcity to accumulate extreme wealth and power

1.2 - Economics

Modern civilization grows and changes as a whole by internal action, because there aren't any outside agents to take purposeful action. There are outside forces of nature, and unintended side-effects which we cause, but those don't involve actions driven by purpose. Economics studies the production, distribution, and consumption of goods and services, and how economic agents and economies as a whole work. Production systems smaller than civilization as a whole, which are agents of change within them, are then also agents in the economics sense. Understanding the effects of self-changing systems includes understanding their economic impacts.

Sneaky Naming - "Credit is what they have - debt is what you have" - a credit is a positive entry in an account book, a debit is is a negative entry. Credit reports and credit cards record the positive entries of the lenders, not your personal situation. What you have is debts, but since that sounds negative (and it is negative), they have renamed things to make it sound nicer, and thus induce you to borrow more.

Modern Feudalism - Parallels of modern wealth concentration and required payments of average people to those with more money to medieval feudalism. We want to escape from this. Community owned production parallels early town charters.

Smart Tools & Economic Collapse

Humans are the original "smart tools". We are able to copy ourselves and pass on our knowledge. We built other tools using food-powered energy and raw materials from nature. Then we used these tools to improve our lives. Technologies like automation, robotics, software, and artificial intelligence are going into new types of smart tools. These new tools don't need us any more to do their job. They will also be able to copy themselves, pass on their knowledge, and make other tools using energy and materials from nature.

The problem is our economic system is based on trading our labor for money, then trading that money for the other things we need and want. If the owners of the smart tools don't need people to run a business, they will get rid of them as an unnecessary expense. But the people who are unemployed will not be buying the products and services those businesses are selling, so those businesses lose income. The unemployed can't afford rent or mortgage payments, so landlords and lenders also lose income. Governments then lose tax sources based on income, employment, sales, and property. If replacement of people by smart tools becomes widespread, everyone loses. It leads to recession, depression, or economic collapse of our current system.

Our tools are only going to get smarter and cheaper with time, but people have a limited ability improve their skills or learn new ones. So eventually all conventional jobs are at risk from this problem. The question is when, and how do we deal with it?

One solution is distributed ownership of the smart tools. People then gain the benefits of what they make directly, without needing jobs and money. Since the smart tools are able to copy themselves, You only need to start with one set of them. Eventually you will have enough sets for everyone. The cost of that first set can then be distributed among all the people who will eventually benefit.

Smart Tools

That's why I said UBI is not sustainable on a large scale. Social programs like welfare only can support a small fraction of the population, while automation is predicted to displace half of all jobs in the next few decades.

The smart tools don't have to be government owned. A government can supply a starter set of tools via a sponsored loan program, similar to how FHA guaranteed home loans work. Since the tools can be used to make more tools, a copy of the starter set can be returned to pay off the loaned equipment (in kind, rather than money). People will then own their own tools, and can use them to satisfy their needs themselves.

[TO BE MERGED]

Eventually, I think "jobs". "prices" and "consumers" will have to be rethought. If corporations get rid of many of the workers, who is going to buy their products?

The way I see going forward is people using smart tools (automation, robotics, software, and AI) for themselves, through cooperatives. By cooperatives, I mean organizations like my power company and credit union, who are both owned by their members. In turn, they hire specialists who know how to run their respective operations.

So, for example, a cooperative "robofarm" might supply food to the members, while hired staff take care of robot maintenance and knowing when to plant and harvest. You skip the intermediate steps of stores and prices. The food goes directly to the members, because they are the owners.

How this becomes affordable is an automated factory can produce the robots and machinery for another automated factory. So the first cooperative builds that first factory. Then enough copies are made to supply all the needs of all the members. If you split the cost of the first factory among enough people, the cost is low per person.

1.3 - Politics & Advocacy

Convincing people and organizations that self-changing systems are coming, and that the technology is worth pursuing, involves some combination of economics, politics, and persuasion.

A political solution: "Re-employment guarantee" - Government jobs supplied to anyone who needs them, and can't find other work, doing useful tasks. But also includes training and start-up support so they can become more employable or work for themselves. This would partly replace welfare and unemployment insurance. Pay would be slightly less than market rates to encourage people not to stay in these jobs forever.

Persuasion - Note that loaded terms like "extreme" affect how people view things, and fear of loss is twice as strong an emotion than opportunity for gain. example: "Do you want to lose your job to robots", of course not. But since robots are coming anyway, it's better to own them yourself than some faceless corporation.

Persuasion through story-telling:

- A story has a protagonist(s) who have a problem which they have to overcome, with opponent/s and roadblocks to reaching a final resolution.

- Our protagonists are several regular people - My father, the trash pick-up guy, a local teacher, and my insuarance agent. Their problem is job insecurity caused by automation, robotics, AI, and software, respectively. Collectively they are beset by these agents of change known as "smart tools" directed by the "evil mastermind" of the soulless capitalist system, whose heart is blackened by the profit motive (greed). Our heroes win by joining together, and bringing the agents of change over to their side, building a new community, and replacing greed with compassion and hope.

Freedom of Knowledge - Knowledge should not be restricted for personal gain.

Dirty air, dirty water, and dirty politics - conservatives and wanting to go backwards instead of making progress.

1.4 - Descriptions, Ideas, & Names

1.4.1 - The Bootstrapping Process

The story of civilization is one of using the TREK principle and Leveling Up.

The TREK Principle:

All of civilization has been built by using the tools, resources, energy, and knowledge we already had to make more of them. This started with our own bodies, sticks and rocks found in nature, the food energy we could gather, and the knowledge that had been accumulated and passed on in the community. As more knowledge accumulated, we could make better stone tools, better shelters for protection, harness the energy of fire, produce more food through agriculture, and develop language and writing to better pass on knowledge. This cycle continues to the present day. We can adopt the "TREK Principle": consciously applying Tools, Resources, Energy, and Knowledge to make more of each. The pieces are still available, all around us. We just have to put them together to bootstrap a better future."

Tools are all the artifacts of civilization used to achieve a desired goal. The first tool is our own bodies, and we still use them for some tasks. But the human body has a limited range of abilities, so we supplement them with a vast range of living and artificial tools. These range from the most basic hand tools up to the most advanced machines. What we generally refer to as tools are the things used to make or maintain other items, as opposed to finished items like artwork that are not. But items like roads and buildings, and furniture like workbenches and chairs to sit at them, are necessary artifacts to make and use the working tools we usually think of. So we include them in the wider class of artifacts used in the production of other items.

The tools themselves plus the knowledge of how to make and use them make up "technology". That knowledge can be internal, as active skills and experience of people, or external as books, journals, data, plans, and instructions. Resources are the inputs used to make new items. They include raw materials, finished material inventory, parts, and already made tools and machines supplied from outside. It also includes less tangible inputs like ownership rights and money. Energy is the ability to do work, both animate an inanimate. It includes various forms of stored energy like fuels and food, sources like sunlight and wind that can be collected, and flows like electric power which can be applied. It also includes available human labor, which is our ability to complete tasks using stored food energy. Living things must frequently replenish their energy, ultimately from inanimate sources like sunlight.

"Leveling Up":

Leveling up is a phrase that originated in role-playing games, and is now common in video gaming. A given character starts at a low level, with modest ability to complete tasks and reach their goals. Over time they can accumulate items and experience, and gain increased abilities. In a broad sense, our entire civilization has been leveling up since it started, with new levels defined by new kinds of tools and technologies: Stone tools, fire, language, agriculture, cities, metals, and so on. We can follow this kind of progression on purpose with our own tools, resources, energy, and knowledge:

Tools - You can start with the first tool, your own body, and learn to use it better. Then you can add manual hand tools, powered hand tools, larger stationary and mobile power tools, and finally "smart tools". These use technologies like automation, robotics, software, and artificial intelligence. As you level up your tools, you can get more done, and do it more easily.

Resources - Tools need inputs to be worked on - raw and finished materials, parts, other equipment to be assembled and maintained, etc. Basic resources can be found all around us for free or at low cost.

Energy -

Knowledge - You can improve your current skills and experience, or learn new ones. You can also accumulate or access external knowledge like books, plans, and instructions. One way to do this is to help accumulate a portable reference library of useful information, which can be shared with anyone who needs it.

You can also choose not to do any of these things and not level up. But then you can't expect your life to change or get better.

You can upgrade your job, or start working on the side. You can also arrange your housing (which is also a tool) to cost less or allow more work space. These improve your finances to start saving money. Then you can apply that money to get better tools, or improve your knowledge, and thus earn more. You can also invest it so it builds on itself. You can then level up your wealth and standard of living. Games teach us that groups with different skills and resources can often work together to reach goals they can't as individuals. In games they are called clans or guilds. In real life we call them partnerships or companies.

1.4.2 - Goals

Continuous Development by Recursive Self-Improvement

In software, recursion is when a function calls itself. In other words, a process is used repeatedly in a loop. More generally, in systems of all kinds, recursive self-improvement is feeding back some or all of the outputs of the system as inputs or mechanisms to make the system better. So a woodworking shop may use its own tools to make a new workbench, which in turn is used to help make further shop improvements. If properly planned, a community, economy, or region can continuously develop by applying this process.

This process isn't new. All of civilization was bootstrapped from rocks and sticks to what we have today. But we can use a more advanced starting point, such as using automated production equipment. We can also intentionally plan an efficient path for self-improvement and development. We all want a better future for ourselves and society. Our vision is to apply the recursion process to reach our goals.

Automated Production - Automated production that bootstraps from a starter set by making more of itself from stored designs.

Self-Replicating Networks: Beyond a single Makerspace to the MakerNet

The peasant farmer can grow his own food, but he can’t make the metal or wooden parts of his tools and home. Conversely, the blacksmith and woodworker can make the metal and wooden parts for the farmer and each other, but not grow their food. All three have to participate in a network to supply each other with what they need. Similarly, in the modern world, a single tool or machine can’t replicate itself, regardless if it is a simple hand tool or a smart robot. A network of people and their tools, supplying each other, can.

Techno-Emancipation, or Freedom From Work

Automated production makes the basics you need to live, such as food, shelter, and utilities. It also makes parts for its own maintenance and expansion. This leaves you free to work on what you choose to do, rather than what you have to do to live. You can use the extra time to work for extras, self-improvement, or just relax, it's up to you.

Futuremakers

I believe in the power of people working together to build a better future(BABF). Our goal is a better future for everyone.

"Planting the seeds of a better future",

"Putting the pieces in place for ABF", or

"Building the foundations for ABF"

"Wouldn't you like to BABF for yourself, your family, for everyone?"

"Use automation ourselves to BABF"

"Helping people to lift themselves up, build a better life, and a better future."

Harness tools, resources, energy, and knowledge to BABF. There's enough of these for everyone in the world to live well. But people are too used to thinking in terms of scarcity. Partly that's because some people create scarcity, and sell the heck out of it for their own benefit.

Factories that Imitate Life - They grow, and evolve, and copy themselves. We want to apply the same powers of growth and reproduction that life uses to endlessly sustain itself. We can copy nature, and make our tools and industry grow and reproduce too, sustaining us and themselves forever.

[TO BE MERGED]

+Steven Sandoval That's pretty much the goal of the Seed Factory Project: http://en.wikibooks.org/wiki/Seed_Factories The first generation design would be used in temperate climates near existing population, and use part of the starter kit (Seed) output to expand the factory, while the remainder of the output is useful items (food, shelter, utilities). Later generation designs would move to remote locations and more extreme climates, and eventually to space. A starter kit isn't able to copy itself right away, but as it grows and adds more machines it is able to make a larger variety of products. End outputs would be things like farm robots, greenhouses, bricks and concrete, and surplus power to supply to people. The application to space is that we cannot afford to deliver whole industrial plants to orbit. Therefore we want to cut that down to just a starter kit, and use local resources to build the rest. The Earth versions will teach us a lot about how to design such self-expanding factories. Although the specific conditions in space will be different, the design process should be similar. — I started writing a textbook for the next generation of space systems engineers. It includes sections on space resources and mining. It's also open source. Who am I? Dani Eder, Recently retired from a career with Boeing's space systems division. Worked on many projects and design studies over the years, most notably the Space Station. Now I want to pass on my knowledge and ideas to the next generation.

Oct 21, 2013, 12:22:34 AM

No, I have a more holistic view. 80% of the Earth are difficult environments (oceans, deserts, and ice caps). I would like to see civilization spread to locations like that, the asteroids, and other planets. The same technologies to use local raw materials and energy apply everywhere, there is no reason to limit ourselves to just some of them. Now, you can argue a most efficient path, and which places to do first, but I would rather not exclude anywhere just because some people don't like it, or it's a bit harder to reach. — I started writing a textbook for the next generation of space systems engineers. It includes sections on space resources and mining. It's also open source. Who am I? Dani Eder, Recently retired from a career with Boeing's space systems division. Worked on many projects and design studies over the years, most notably the Space Station. Now I want to pass on my knowledge and ideas to the next generation.

Jun 4, 2013, 9:06:43 PM

The underlying problem is corporations work for their shareholders, and therefore screw the workers and customers. The solution is for people to work for themselves through community cooperatives, like my power company and credit union. They are non-profit and member-owned and we get a better deal all around.

In the 1930's the US backed "rural electric cooperatives", to bring electricity to the rural areas that for-profit power companies wouldn't serve. The co-op members did the work themselves, but the government provided technical help and start-up loans to get them going.

We could follow that model and set up business co-ops, especially in rural areas where good jobs are lacking. Have them build and manufacture stuff again. As owners, the workers aren't going to screw themselves, and pay fairer wages. The people at the top can still earn more, but not 300 times more, and they won't skim off profits for the shareholders.

1.4.3 - Names and Slogans

Sorted alphabetically to reduce duplication:

@Home Movement - Factory, farm, banking, etc literally at home or locally.

Artificial Systems which become something else over time
Automated Production that Evolves
Automated Replication
Automation Village - A demonstration community using high levels of automation. A community owned business
Autoseeds - Another name for the starter kits for self-expanding automated factories.

Bootstrapped Replicators - From dumb to smart tools.
Bootstrapping Technology - Methods to upgrade and expand from starter sets to higher levels of technology.

Civilization in a Box - Deliver various sized "boxed" starter sets, with different grades and amounts of TREK elements (tools, resources, etc.)
Fabrication Network - A name for a distributed production system where different people own some machines in a production node, and collaborate over a network.
Growth and Replication Tech - Technology that copies what living things do.
Human-Augmented Replicators
Life Improvement - by analogy to "home improvement", but applying to all aspects of life.
Production systems that self-expand from a starter set

Replicating Production Systems - A production system makes stuff. A replicating production system makes itself too.
Replicator Seed - Another description for a Seed Factory.

The Seed Factory Project - "After 10,000 years of civilization, isn't time to end scarcity?"
Seed Tech - A short description for the concept.

Self-Bootstrapping Automation
Self-Changing/Modifying Systems
Self-Expanding Systems
Self-Improving Networks
Self-Modifying Systems
Self-Replicating Communities
Self-Replicating Networks
Self-Reproducing Distributed Factories
Self-Supporting, Sustainable Communities with a High Quality of Life

Toolbox Evolution - Self-expanding, growing, modifying, changing, upgrading, improving.
TRUE Growth System - Tool Replication, Upgrade, and Expansion.

Viral Factories - that can copy themselves

2.0 - Starter Locations & Network

These projects begin gaining experience building and operating starter sets, bootstrapping, and self-expanding systems. They are in moderate environments, in already developed and populated areas, which are the easiest places to start.

2.1 - Personal Production

Personal Production Network

A personal production network is made up of a distributed group of people who work together and help each other.

- They begin with a starter set of skills, knowledge, tools, and resources.

- They bootstrap from there by building and operating their own equipment, and creating useful products and services for themselves and others.

- They help each other upgrade and expand their capabilities. This includes:

-- Learning new or improving existing skills,

-- Adding more information, designs, and plans to their knowledge base,

-- Adding more tools, equipment, and workspaces to the network by purchase or making their own using what they already have, and

-- Accumulating funding, materials, energy, parts, and other resources to work with.

- Their goals include:

-- Meeting most or all of their personal needs through the network,

-- Becoming self-supporting and economically secure, and

-- Able to help new people by growing the network or starting new ones.

A fully developed network would be able to make everything members want, including copying all their equipment and supplying new starter sets. A self-contained production chain, from raw materials to finished products, makes self-expansion and upgrade easier and less expensive. "Smart tools", which exploit automation, robotics, software, and artificial intelligence, can grow efficiently with little labor. A new network won't start out with all these abilities, but would grow into them a step at a time. As owners of the smart tools, they are not at risk from labor displacement. They still benefit from the tools, no matter how smart they get.

Business Venture Fund - This is a way to fund development of the network. It fills the "accumulating funding" item above. Venture members pool their savings and first invest in conventional income-producing investments. Once they have enough, members can start buying equipment, land, etc and start up businesses and production. This is more than just pure investment, it is investment with a purpose.

2.1.1 - Names and Descriptions

Bootstrap a Better Future: by networking and making our own stuff.
Bootstrap Co-op
Building our Future Network

Community Automation
Community Bootstrap Network - Bootstrap our own ability to make things via increasingly smart tools.
Community Development Co-op: - Improve our homes and ourselves
Community Exchange - similar to Community Exchange System but emphasis on tools, making, and building.
Community Project Network
Community Upgrade Co-op
Community Upgrade Network

Co-Working Network
Extra Hands Project Network

Life Improvement Network - Making life better in all kinds of ways: self-improvement, training, help & service to each other, home & real estate upgrade & development, ootstrapping from smaller projects, etc.

Make & Build Network - Neighborhood network for helping each other get things done. A network of people who can make things for themselves and each other.
(Atlanta) Southside Improvement Network - Build a better life for ourselves & each other

Neighborhood Project Network

Production & training incubator center

Phase 1 Names: "A Project Network to build a better future", Home Co-op,

Seed Factory Technology Center
Seed Factory Network
Self-Help Network - Help people to help themselves, because corporations sure won't once automation allows it. We need to learn to do things on our own.

Phrases

The difference between working together and working for someone else is you get to keep all of what you make.
Grow the ability to make anything we want on command, using smart tools. Voice command already exists (Alexa & others). Computer-controlled factory machines and robots also exist.

2.2 - MakerNet

A network of people, and their tools, resources, and knowledge. They build real wealth and security by making useful products, with a goal of meeting all their basic needs and replicating their tools to expand the network.

Features

Programmable Self-Expanding Factory - Uses combination of direct labor, remote operation, automation, and robotics.
Mobile Factory Elements - Bring the factory to the worksite.
Versions: - Residential, Industrial, Distributed, Difficult/Remote, Space.
Growing System - Containerized hydroponics using waste cardboard boxes, organic mulch (like leaves and sawdust with compost) and a manure tea.
Robotic Proxy - Using a tele-operated robot to observe or perform tasks. Reasons include reduced commuting, living in preferred locations, or the task location is remote or hazardous.
Countering Resource Depletion - Premium resources (e.g. high grade ores or easy to mine) are limited. So by recycling and extracting from lower grade sources we can extend the life of all resources.
Self-Sustaining Communities - That can efficiently make most of their own physical needs via automation, and also do other/outside work for services and fill in what they can't make.

Design

Resource Model - For spreadsheet version, use separate lines for alternatives, with % weight column to switch or mix choices. Use separate pages for phases, with flows between them for growth, and different table values and added lines to indicate state of the factory. Flow diagrams can also be included. Research free modeling tools besides spreadsheets.
PV Tracking Mount - To increase output from purchased PV arrays. Since we cannot easily make our own solar panels from scratch, we can supplement their output with an active mount. We use single polar-axis mount and ignore seasonal tilt because the concentration ratio is low. Use side reflectors to increase collection area. Make sure panels are qualified for higher temperature use.
Makerspace/Community Workshop - Build excess space and lease out use to help with growth. Possible roll-away lockers, with lockable carts or cabinets, so that people can put away their work, and easily pull it out and move it to a work area when needed. Larger sizes of storage/private work areas, and a merchant/vendor area (like indoor flea markets). Ample parking.

Automation

The Paradox of Automation - Corporations are legal entities, they don't design and build anything, people do. Thus whatever technology is developed within corporations will also be available to individuals.

The Automation Revolution - Parallel to the Industrial Revolution. The latter replaced much of human labor with new power sources. Automation will replace much of human thinking with new computation resources.

3.0 - Commercial & Industrial Development

Expandable Factories - As self-expanding systems grow, they go beyond personal scale, to commercial and industrial scale.

Notes from "Engineering Mega Systems" (through p55/257) - In the transition from an industrial age to an information age, so factories will be based more on information than steel. Thus the network and shared information is becoming more important. Hardware protocols may be useful by analogy to network protocols for data. Hardware "packets" flow through the factory, and machine interfaces and networks are defined.

Applying Metcalf's Law to factories: Number of products and usefulness goes as square of number of tools and production elements?

3.1 - Production Centers

A shopping center is where you go to buy things that are already made. A production center is where you go to get things made. The rise of automation and computers makes it possible to custom-make items on demand. Grouping design, fabrication, and supply/warehousing makes it more efficient to do this, and promotes customer traffic. In cases it makes sense, a production center can support mobile operations (i.e. construction and home improvements), distributed production (marketing, supplies), provide startup space and support, and training for people who want to work or start new businesses.

4.0 - Developing Earth's Frontiers

These systems expand into difficult and extreme environments. They are supplied at first from previous systems in easier environments, then start to trade with the growing network.

Earth Colonies

The world seems a crowded place to many people, but this is a result of selection bias. Most of us live in places where other people live, so that is what we see around us. For example, in 2016 55% lives in urban areas with a lot of development. But this only makes up about 1% of the land area or 0.3% of the total planet's surface. The remaining population is distributed more widely, but only about 13.5% of the Earth's surface is used to any great degree. This includes urban, forest, and farm lands. The remainder is areas like oceans (70.8%), deserts (3.5%), ice sheets (3.1%), and tundra (2.3%), which are hardly occupied and lightly used. Even the part we use amounts to only a thin surface layer on average. If we add up the entire biosphere, of which we use a significant fraction, plus everything humans have made, and distributed it evenly over the 13.5% of the Earth we use, it would make a layer about 18.5 cm thick, about a hand-span. In a very real sense we have barely colonized our own planet. So how can we talk about colonizing Mars or other places in space?

The answer is that any given location, on Earth or in space, is easier or harder to live in based on the local environment and resources. We are an adaptable species, but we tend to occupy the easiest places first. As our technology improves, we can occupy more difficult places and still live well. For example, in the US, the development of air-conditioning allowed a large population shift to the hotter south part of the country. Heating by means of fire is an old technology, so we were better able to live in cold places, but cooling by means of electric devices is relatively recent. Cost also matters. At any given time, quite a few people are living aboard cruise ships, but most of them are tourists who only stay a few days due to the high cost. Relatively few people live on ships if their job isn't there, due to the great expense involved. Thus we don't yet have floating cities on the oceans. But if better technology brings the cost down, such cities may evolve.

Colonizing harder places on Earth would be good practice for colonizing space, because you have to solve the same kinds of problems. No matter where you live, you need protection from the outside environment, food, and basic utilities like water and electricity. If you want to be self-supporting, rather than depending on trade, you also need to make things using local energy and materials. The farther you are from other developed areas and the greater the difficulty of transportation, the more important it becomes to use local sources.

5.0 - Developing Space

Development is meant in the same sense as public infrastructure and real estate development on Earth - improvements from the natural state.

Space Systems Engineering Book - Updated names - Future Space Systems & Programs, Future Systems for Earth and Space, Handbook of Future Space Systems

5.1 - Space-Related Ideas

Analyze "shadowing" (availability of solar energy) as a function of geography (latitude, altitude, & topography) and orbits, to identify preferred operating locations and need for alternate power sources or storage.

Economic Development of the Solar System: A Vision for the Future (article)

We treat economic development as a process which should be extended to the entire solar system, including the large parts of the Earth not currently used. We discuss a roadmap for this development starting from what exists today, and leveraging smart tools and self-expanding systems.

Using the BFR to Colonize Mars

SpaceX has said their nominal goal is to colonize Mars by putting a million people on the planet. They have also said the nominal capacity of their BFR is 100 passengers to Mars. It therefore takes 10,000 trips to Mars to deliver a million people. To get to Mars, the ship is refueled in Earth orbit, which takes about 7 tanker launches (1100 tons propellant @ 150 tons/tanker). Therefore the total number of Earth launches needed is 80,000. This ignores the cargo needed to support the passengers once they are on Mars, which the SpaceX paper suggests is two cargo for each passenger, on at least the first few flights. A 2:1 ratio of cargo to passengers then results in 240,000 total Earth launches. With a nominal BFR cost of $10 million/flight, total transport cost comes to $2.4 trillion.

The total launch count and costs are high enough to make it worth looking at more efficient methods. The BFR is probably adequate for the first 10,000 people, enough to establish a viable colony. This requires ~2400 launches at a cost of $24 billion. Assuming 1000 people are delivered per Mars transport window, which happens every two years, you would need 10 BFS passenger and 20 BFS cargo upper stages, and sufficient tankers and launch pads for a total of 240 launches per window. Launch rate would depend on how long the upper stages can loiter in orbit while being refueled. In the limit of the full two years, you would need 120 launches per year. These seem like feasible numbers. For the full million people, assuming 100 years to deliver them, requires 20 times the launch rate, at which point serious consideration should be given to better approaches than chemical rockets.

Space Elevators

Space elevators are a real engineering concept. I've worked on them for Boeing and NASA. However the simple version that gets all the media attention - a 60,000 km cable from ground to beyond synchronous orbit - isn't possible with known materials. More advanced versions which take advantage of orbital mechanics and use shorter cables are quite possible with today's materials.

However, a space elevator is "transportation infrastructure" like an airport or a bridge. They are expensive to build, but cheap to use for each trip. Therefore you need many trips for them to make economic sense. Total Earth-to-space traffic isn't big enough yet for *any* space elevator to pay for itself. Research for the future is still worthwhile.

Seed Factories for Mars

Learn about "seed factory technology", or as Musk calls it "the machine that builds the machine". Whoever knows about this stuff will either be hired by or become subcontractors to SpaceX.

In order to support millions of people on Mars, you are going to need a lot of industry: agriculture, mining, construction, etc. There is no way you can carry all the equipment needed for that from Earth, even with regular BFR flights. The answer is to build most of it from local materials and energy, using a starter set of equipment to build the rest. That's the "seed factory".

It would have things like solar furnaces, machine tools (lathes, milling machines, etc.), generic electric vehicles with attachments the way farm tractors work, etc. Mars has [metallic meteorites](https://photojournal.jpl.nasa.gov/jpeg/PIA18387.jpg) just sitting around on the surface. They came from the Asteroid Belt, which Mars skims the inner edge of. There are probably more of them buried under the soil, but you can find them with a metal detector or ground-penetrating radar. They can be your raw materials for metals production, and then later machines would work other materials.

Vacuum Distillation

Water is available in Chondrite type asteroids, and if the extraction process needs some, you can get it. More likely you would use vacuum distillation though. Concentrated sunlight is easily available in space. So you evaporate chunks of metallic asteroid by heating. In a vacuum, evaporation begins before melting. Then you arrange a condenser with varying temperature along the length. Chosen metals will plate out at their respective condensation points.

Transfer Habitat

I have quite a different conception of space development and colonization than either NASA or SpaceX is have. You can find a short version in my report To Mars and Beyond, and I am developing the ideas in more detail in part 4 of the Space Systems textbook I have been working on (it is not finished yet).

I feel that both NASA and SpaceX have an excessive focus on rocketry, and not enough on using the material and energy resources available in space. The inner Solar System has tens of thousands of asteroids. It is not a void to be gotten through as quickly as possible on the way to Mars, but rather filled with raw materials and flooded with energy by the Sun.

So if you really want to put a million people on Mars, as Elon Musk has said he wants to do, we should use the available resources to do it more efficiently and safely than shuttling a "Big Effing Rocket" back and forth repeatedly. My approach is to mine and process materials at multiple places on the way to Mars. This generates propellants and other useful products along the way and greatly reduces what you need to bring from Earth.

The middle part of the journey would include one or more "transfer habitats" in permanent cycling orbits between Earth and Mars. Mining tugs would collect raw materials from "nearby" (in delta-V terms) asteroids. Because there are so many asteroids, you are guaranteed to have some that are nearby in this sense just from statistics. The raw rock can provide radiation shielding, and a source of propellants, water, and other useful products. Doing the processing gives the crew something useful to do on the trip, and Mars colonists will have the necessary skills, since they will be doing similar work to live on the planet once they arrive.

Because the habitat makes multiple trips, and is supplied from local resources, the total mass/colonist launched from Earth is drastically reduced, and you don't need a giant rocket like the ITS. You certainly don't need to accelerate the life support for the passengers from Earth to Mars and back again each time. Rather, you have smaller crew capsules which accelerate at each end of the trip, and the rest of the time they are in the larger and better shielded habitat.

The extraction of propellants and supplies doesn't only happen on the transfer habitat. You do it at Earth-Moon L2 and probably Phobos. We have yet to visit Phobos and don't know for sure if it the right composition. If not, Mars skims the inner Asteroid Belt, and there are plenty of other choices there. At EML-2 you produce propellants to inject to Mars transfer, and for other destinations like the Moon and lower Earth orbits. The mass return for a mining tug is about 200:1 relative to the starting mass that comes from Earth, so it provides great leverage.

You also develop mining and a catapult on the Moon. The mass return from the Moon is even higher than from asteroids, but due to the Moon's history, it is depleted in volatile compounds. To make a wide range of products, you want both sources. You want to do the processing at EML2 because it gets sunlight 100% of the time, while the Lunar surface only gets it half the time. You can also choose the gravity level in orbit according to your needs. Zero gravity makes large assembly easy, while other processes and people need some gravity, and with rotation you can get whatever you need.

The transfer habitat can also be under rotation, and with radiation shielding there is no need to rush the trip. It can also be much more roomy than the ITS passenger module, have a greenhouse for fresh food, and large solar arrays for life support and comfort features. It would be more like a mobile colony than living in an airplane for four months. Building such a large habitat is justified by the fact that you intend to transport a million people over the course of a century.

Mars Skyhook

If we are seriously going to colonize Mars, we would build a two-segment space elevator system + ground catapult to efficiently get up and down from the surface to desired orbits. In the long run it would be much more efficient than rockets, and a two-segment design is much easier to build than a one-piece elevator you see in popular articles.

The numbers work like this:

- You build an 85 km electromagnetic catapult on Pavonis Mons, a Martian volcano that sits on the equator. It has ~120 of slope on the west side, so there is room for it. The top of the mountain is 14 km above Mars' reference altitude, so air pressure is about 180 Pascals, or 0.175% of Earth's. Therefore air drag should be minor once you leave the catapult. At 30 m/s^2 acceleration (3 g's), which is low enough for people, you reach 2265 m/s at the end of the catapult.

- You build a 175 km long (87.5 km radius) "skyhook" (or rotovator) type elevator in a 240 km low Mars orbit. It rotates end-over end with a tip velocity of 926 m/s. Accounting for Mars' rotation at the equator, this is just enough at the low point to match the velocity of payloads thrown by the catapult. At the upper end of rotation, 926 m/s is the amount needed for Phobos transfer orbit (this is what sized the skyhook). The tip provides 1.0 g's acceleration, making it comfortable for people.

- The total stress on the cable is 0.5 x tip acceleration x length = 43.75 g-km. Modern carbon fiber has a safe working stress of 150 g-km, so we are well below the limits for this material. Working stress, in turn, is 2.4 times below ultimate stress, because we never design structures to the point of failure. We always have a margin of safety. Phobos is likely a carbon-bearing asteroid. We won't know for sure until we send a probe there, but spectral readings from a distance indicate it is. So we can probably get the carbon we need from there.. If not, we can get it from a nearby asteroid or the Mars Atmosphere, which is mostly CO2. Mars skims the inner edge of the Asteroid Belt, so there are *lots* asteroids to choose from.

- By choosing at what distance from the center of the skyhook you let go, you can reach any intermediate orbit you want up to Phobos transfer. A second skyhook at Phobos of a similar size can pick you up from transfer orbit and send you on to Mars escape, or any intermediate higher orbit. So this system in theory allows you to get up and down from Mars without using any propellant. In practice you will need a small amount for adjusting arrival velocity to match and variations in Mars' atmosphere.

- Two smaller skyhooks would be much easier to build than a 12,000 km long one-piece elevator anchored to Phobos. They would also have much less exposure to meteorite damage, which is a significant worry when you are near the Asteroid Belt's edge. Both designs would need multiple cable strands to withstand impacts, because you can't detect centimeter-sized meteorites coming in. However, the total damage rate is proportional to exposed length, so shorter is better, by a factor of 35 in this case. The shorter skyhooks also leave the space between them open for other uses.

Cost to Orbit

Aerospace systems generally run about $2000/kg. For example the Boeing 737-700 costs $85.8 million and has an empty weight of 38,140 kg, giving $2250/kg. Rockets are simpler than airplanes (no cockpit, passenger seats, etc.), but they are also produced in smaller numbers, so let's round down to $2000/kg.

The Falcon 9 has an approximate empty weight of 31,700 kg, giving a cost of $63.4 million for the hardware, and the listed price is indeed $62 million. RP-1 (a type of kerosine used in the Falcon 9) goes for about $1285/ton, and Liquid Oxygen is only $66/ton (there is a plant that makes it from air at KSC, so it is really cheap). Total mass of the rocket is 549 tons, and payload is 22.8, so propellant is 494.5 tons. The mixture ratio is about 2.5:1, so you use 141 tons of RP-1 and 353 tons of Oxygen. Total propellant cost is then $204,483, which is in accord with Musk's quote that it costs $200,000 to fill the rocket.

So you can see that by far most of the cost is the hardware. Propellant only gets used once, and past rockets used the hardware only once too. Using it multiple times saves a lot of money.

This is why you can afford to ride an airplane and not a rocket. Airplanes fly on the order of 20,000 times, so that 737 costs $4290/flight for the hardware, or $25 per seat. Musk's goal with the Falcon 9 Block 5 is to fly it ten times, and for the Big F**cking Rocket (BFR) more than 100 times. The propellant to fill the rocket will still be about the same cost, but the hardware cost will be much lower, bringing the total cost to fly down a lot.

I have a physics degree, and have worked on Space Systems Engineering (i.e. "rocket science") for 40 years now. The ultimate cost is set by the cost of energy. The amount of energy is set by the Earth's mass and radius. Wholesale electricity is ~$0.05/kWh, and the energy to reach low orbit is 33.2 MJ/kg = 9.23 kWh. Therefore the theoretical cost at perfect efficiency is $0.46/kg, less than Walmart sells bulk flour for.

However, chemical rockets are not perfectly efficient. The fuel energy of the Falcon design is 12 MJ/kg, and it is 90% fuel at liftoff. So sitting on the pad the energy content of the rocket is 10.8 MJ/kg. The 4.15% by mass of payload acquires the 33.2 MJ/kg of orbital energy, or 1.38 MJ/kg of the total rocket mass. So the efficiency is 1.38/10.8 = 12.77%. Therefore the lowest possible cost of a chemical rocket is $3.60/kg.

Obviously real rockets have other costs besides purely energy, but this is a floor set by physics. Other transport methods besides chemical rockets are not limited by their low efficiency, and theoretically can get closer to the cost of raw energy. Usually these other methods have high design or infrastructure capital costs, which have made them uneconomic to pursue. Assuming cheaper chemical rockets expand the market enough, those other methods now become worthwhile. So in the future we may go beyond the inefficient rocket approach, but we must first go through that stage of technology (pun intended) to get to them. My textbook that I linked to talks about what those future options are.

Numbers for Space Mining Requirements

World Minerals Production 2015 = 17.27 billion tons

Construction Aggregates 2014 = 40.2 billion tons

Water use per capita = 506 tons/year (69% agriculture, 19% industry, 12% household & municipal)

ISS consumables/person/day:

Oxygen 0.84 kg

Drinking Water 1.62 kg

Dried Food 1.77 kg

Water for Food 0.80 kg

Total 5.03 kg = 1.836 t/yr

Passenger airplane volume = 2.6 m^3/person

BFR = 8 m^3/person

Skyhook Systems

As someone who has taught a [space elevator](https://en.wikipedia.org/wiki/Space_elevator) class, this is a poorly written article.

First, there are now [carbon fibers](http://www.torayca.com/en/download/pdf/torayca_t1100g.pdf) (399 km) with higher strength/density than Zylon (384 km). But pure strength isn't the only reason you choose a material.

Second, there are much better designs than the Tsiolkovsky 1895 one (ground to synchoronous orbit in a single cable). Hans Moravec's 1977 Skyhook concept is much more efficient. For example, a skyhook orbiting at 7500 km radius (1122 km altitude) moves at 7290 m/s. If it has 1000 km long arms, so as to miss the atmosphere at the lowest point, and rotates at 2 g's at the tip, the rotation velocity is 4472 m/s. Since the Earth rotates at 465 m/s at the equator, the tip moves at 2353 m/s relative to the ground. This is 1/3 of orbit velocity, and easily reached with a conventional rocket. Such a skyhook has a mass ratio of 820, vs 10 trillion for the Tsiolkovsky design.

Third, Engineer Zero plugged in the *ultimate breaking strength* into the space elevator formula to get the million to one taper ratio. This is wrong, because real designs don't stress materials to their breaking point if you want them to last. With a reasonable factor of safety vs breaking stress of 2.4, you get the ten trillion value.

You can do even better with a three-part system rather than a single skyhook. This would include a ground accelerator that provides 1550 m/s, a lower skyhook that supplies 3000 m/s at 1.5 g's and one in high orbit. The lower one has a radius of 600 km and a mass ratio of 30. The rocket needs to supply:

7504 m/s orbit velocity - 3000 m/s skyhook rotation - 465 m/s earth rotation - 1550 m/s ground accelerator = 2489 m/s, about the same as before.

Warming Mars

Heck, there's an estimated 5 million cubic km of water on Mars, and water is a good greenhouse gas. We just need to get some of it vaporized to warm the planet up. An orbiting satellite beaming microwaves at 2.45 GHz should work (same as a microwave oven)

Orbital mirrors are possibly the easiest approach to warming Mars. Assume you want to double the solar flux, to bring it up to about Earth levels. Mars' cross-section is 36 million km^2, and the reflectivity of aluminum is about 85% for sunlight. So you need 42.7 million km^2 of mirrors. Assuming your reflectors are 25 micron aluminized Kapton (a standard product on Earth and used on satellites a lot), it would be 25 cubic meters/km^2, or 1.068 billion cubic meters total. This is 1.068 cubic km. The density is 1.42, thus just over 1.5 billion tons total. By comparison, the current Mars atmosphere masses 25,000 billion tons. So any significant change to it involves a lot more mass than the mirrors.

At Mars perihelion, the Sun subtends 0.0067 radians in width. A flat mirror will project an image of the Sun of the same angular size. In order for all the light to hit the planet, the mirror needs to be orbiting closer than the distance for which Mars is the same width, or anywhere within about 1 million km. If you can make it so the solar light pressure equals Mars' gravity at some distance, the mirror can be stationary relative to Mars, which makes operations simpler. Otherwise the mirrors will be orbiting and have to aim the reflected light.

Doubling solar flux should raise the planet's temperature from an average of 210 to 250 K (from blackbody formula), but polar vs equatorial, and side effects from evaporating CO2 and water would complicate things. You'd have to do a full planet climate model.

Terraforming Venus

The best way I know to start terraforming Venus is sunshades - lots and lots of sunshades. Venus has a cross-section of 115 million km^2. If your sunshades are 1 micron thick (10^-9 km), in principle they would total 0.115 km^3 of asteroid material, which isn't much for a terraforming project. If they are thicker than that, just multiply by the actual thickness.

You want them flat on the sun-ward side of the planet to block the Sun, but open to space on the other sides, so as to let heat escape. If they are in orbit rather at the Lagrange point, just turn them from flat to edge-wise as they orbit.

Blocking the Sun will cool the planet on a time scale of a century or so. It is a thick hot atmosphere, and the surface below it is also hot. As the atmosphere cools, the scale height will decrease. That means the pressure on the high spots, like Ishtar Terra, will decrease more than the surface. The higher elevations will also be cooler. So those are the first places we can occupy.

If we are lucky, the right kinds of minerals exist on the surface to absorb CO2 via carbonation, for example:

"research at the Albany Research Center (10,13) has focused upon the direct carbonation of olivine. When the program first started, it took 24 hours to reach 40-50% completion of carbonation of olivine. The reaction required temperatures of 150-250 C, pressures of 85-125 bar, and mineral particles in the 75-100 micron size range."

https://www.netl.doe.gov/publications/proceedings/01/carbon_seq/6c1.pdf (page 6)

Those temperatures and pressures are not so different from what exists on Venus. This would lower the atmospheric pressure and make the lowlands more habitable.

Once the planet has sufficiently cooled/absorbed, you can then allow partial sunlight through to stabilize conditions.

Thermal flux from Venus is 160 W/m^2. The atmosphere has a mass of 1.037 million kg/m^2 and is 96.5% CO2. CO2 has a [specific heat](https://www.engineeringtoolbox.com/carbon-dioxide-d_974.html) of roughly 850 J/kg-K. It varies with temperature, and so does the atmosphere with altitude, so I took an average.

To effect 1 K of temperature drop would then take mass x specific heat = 881.6 MJ/m^2. Heat loss in 1 year is 160 J/s x 31556925 s/year = 5.049 MJ/year. So the initial rate of temperature drop would be 5.7 K/year.

Thermal flux is a strong function of temperature, so it won't stay that fast. You also get complications like sulfuric acid raining out once it gets cool enough (the clouds evaporate before reaching the surface today). Since we need to lose 437 K to reach room temperature, we can say the initial cooling has a time constant of 76 years. In other words, if it stayed the same rate (which it won't), it would reach room temperature in that time.

Another complication is the surface of Venus is likely basaltic in composition, and CO2 is driven out by the high temperature. As it cools, some of the CO2 may be re-absorbed into carbonate minerals, lowering the pressure. We would need to know the composition and physical state (dust vs slabs of cooled lava) of the surface to know what would happen.

Propellant Extraction

First, you want to fetch carbonaceous asteroid rock and fetch it back to a high orbit like L2. You then don't have to pay the mass penalty of launching it using chemical propulsion from the Moon, and that type of asteroid has up to 20% water and carbon compounds.

An electric tug of 10 tons hardware mass and 26 tons starting propellant can haul back 1000 tons of rock, yielding up to 200 tons propellant. You use 26 tons to fuel the next trip and send the tug back out. The rest can be used for other missions. Round trips take 2-3 years, depending which asteroids you visit. The tug can reasonably last 15 years before needing new solar arrays and engines, making 6 round trips. Net propellant is then up to 1000 tons over the life of the tug. If you need more, or need it faster, launch more tugs or build bigger ones.

Near-Earth asteroids don't have water as water, they are too close to the Sun for it to survive. The water is in the form of "hydrated minerals", such as serpentine or clay. To get the water and carbon compounds out requires heating to 200-300 C. A solar furnace, pressure vessel, pumps and plumbing, and a condenser are needed for this.

If we are baking 1000 tons of rock every 2.5 years (one trip of the tug), that comes to 1.1 tons per day. The heat capacity of space rocks is on the order of 1 kJ/kg/K. Heating 1.1 tons to 300C thus requires 330 MJ. Divide by seconds in a day yields 3.8 kW. A solar concentrator slightly over 2 meters in diameter would theoretically be enough. Given heat loss and process inefficiencies, increase that to 5 meters. You pass the crushed rock or dust through the focus at the rate of 12.7 grams/second, and tap off the evaporated gases with a vacuum pump.

This is not a 20,000 processing plant. It is more like a few tons. You will need large tanks for the accumulated gases, and condensers are different temperatures to separate out the products.

Solar Power vs AU

Metallized Kapton film is used all the time on satellites. The gold-coated version is good for thermal control, and the one most commonly seen on them, but aluminum-coated has a better reflectivity. A standard thickness plus reinforcing filaments can supply 80 square meters of reflectors per kg. Modern space solar arrays produce 175 W/kg at 30% efficiency, or 408 W/m^2. Therefore their mass is 2.333 kg/m^2.

You can therefore supply 186.65 square meters of reflector area per square meter of solar array and only double the mass. The reflectivity of aluminum is about 85%, but I will assume 75% to allow for non-perfect delivery of the light to the panels. Our net illumination is 140 suns. Since solar flux falls as the square of distance, we get 2 suns illumination at 8.366 AU. Since our mass is double that of the bare panels, the net output is back to the original 175 W/kg.

The "Kilopower" small fission reactor being developed by NASA is expected to produce 10 kW at 226 kg mass, or 44.25 W/kg. The ratio of solar is 3.955 times higher, so we can go 1.989 times farther from the Sun before nuclear is more mass-efficient, or 16.64 AU. This is between Saturn and Uranus.

The Dawn mission, which visited Vesta and now Ceres in the asteroid belt has run just fine off solar, and the Juno mission at Jupiter (5.2 AU) represents the state of the art for solar *without* reflectors. With the addition of reflectors we can push the limit another factor of three before nuclear is the best option.

Note: these calculations are based on current or near-term (in test) technology. We don't have reliable data for large space fission-electric reactors, and we don't know if and when fusion reactors will be available, and what their mass will be. So any discussion of whether they are the preferred option for a given mission is premature.

Power from Space

You are missing the trillion-dollar energy market. There is 4-10 times as much solar energy in space compared to places on Earth, and if you avoid the Earth's shadow, it is daylight all the time. Space industry will eventually produce 98-99% of its products from local materials. Satellites that beam energy to Earth produce about 100W/kg. 1-2% of that is 0.1-0.2 mg from Earth per Watt. At three times BFR launch cost goal ($20M/launch), transportation is 2-4 cents/Watt. Current solar and wind on the ground is $4/average Watt, so launch will be a negligible cost overhead.

What's 24-hour clean energy delivered anywhere on Earth worth?

Obviously, there is a lot of work to do before this is economically viable. First get launch costs down, then bootstrap space mining and industry. Assuming you want a TeraWatt beamed down, that's 10 million tons of satellites, and probably 3-10 times that in raw ores. It's 100,000 to 200,000 tons from Earth, or 1-2,000 BFR launches.

1 TW around the clock is 8.76 billion MWh. One MWh goes for $30-50 these days, so we are talking $260-438 B/yr. Civilzation consumes 20 TW currently, so you don't have to capture the whole energy market.

Why Go Back to the Moon?

> Why didn't we (humans) continue to go on the moon?

Because the "space race" of the 1960's was part of the "dick waving contest" between capitalism and communism, to show which system was better. The US represented the capitalists, and the USSR represented the communists. The undecided countries is who we were trying to win over.

Once the US won the race, the motivation to keep running hard went away. However, closing government offices is very hard. NASA's budget was cut 2/3rds in real terms, but all their centers stayed open, and they have continued to work on projects at a slower pace ever since.

> why is the moon become "hot" again?

Just because the US/USSR race was over, didn't mean activity in space stopped. Today there are ten times as many active satellites in space as in 1969, and those satellites are vastly better, due to improvements in technology. We are now approaching the natural time we *would* have gone to the Moon, if the space race hadn't happened.

There has been an occupied space station in low orbit for the last 20 years. We have learned a lot of things on the Station. But the most important thing we learned was how to assemble large, complicated things in space, and keep them running. We didn't know that in 1969, so the missions only lasted a few days each.

We also now have partly reusable rockets, and soon fully reusable rockets. The Saturn V (and SLS) are entirely thrown away after you use them once. That makes them very expensive. Reusable rockets vastly bring down the cost, so we can afford to do more on the same budget.

The last thing, which has only been tentative and experimental so far, is using material and energy resources already in space for our projects. The more we can use stuff that is already there, the less we have to launch from Earth, and the further we can drop costs.

The Moon is covered to an average depth of 5 meters with loose rocks, sand, and dust. So we don't even have to use explosives or jackhammers to start mining. All we need is a shovel. So it is an easy place to start using off-planet resources.

Off-Planet Mining

A high orbit near the Moon makes sense if you get seriously into off-planet mining and production:

We define "mass return ratio" as mass of mined ore divided by mass of mining equipment. For the Moon this is on the order of 3000:1, and for nearby asteroids it is on the order of 200:1.

This ratio is regardless of the cost of launch from Earth. The big SpaceX rocket will likely still cost ~$20M per launch, which works out to $200/kg, a high number in earthly terms. So if you can leverage it by using off-planet materials, you can still save money in the long run. In the short run, the R&D for the mining and production equipment will cost more.

The Moon has two basic geologies - highlands and maria, and asteroids have three main groups - carbonaceous, stony, and metallic. All five are different from each other due to their different origins and histories. The history of mining tells us to mine where the ore qualities are best. Therefore you want to mine all the sources. By combining sources, you can use a wider range of processes and make a wider range of products.

"Embodied energy" is the total energy used from mining to finished product. For typical Earth products it is in the range of 10-20 MJ/kg, although the range goes much lower and higher for some products. The energy to get raw materials off the Moon to low orbit is 1.5 MJ/kg, and to escape is 2.83 MJ/kg. This is much less than the embodied energy of finished products. Since high orbits get twice as much sunlight as most places on the Moon, your space factories will be more productive in high orbit. Also, most of your customers will be in Earth orbit and other locations not on the Moon.

There are special cases like the Lunar poles, which get continuous sunlight and have cold traps with water. My points above apply to the general case of off-planet mining.

Celestial Billiards

You can play what I call "celestial billiards". There are nearly 20,000 known near-Earth asteroids, most of which are small. Aim one of the small ones at the dangerous one, to knock it off course. Assuming you have a year's notice, you only need to change course by 0.5 meters/second to miss the Earth by 3 sigma (20,000 km). If the "cue ball" has a 5 km/s impact velocity, it can be 1/10,000th the mass.

So, megaton asteroid heading for us, 100 ton "cue ball", which would not take much to change course of. You pick one that's already going to head near the target. Out of your 20,000 candidates, there should always be one.

Interstellar

> Even the most optimistic estimates would give 0.05% light speed for a feasible generation ship

I'm more optimistic than you, and also a rocket scientist. 0.05% x c is 150 km/s. We can already feasibly reach that speed with current technology: nuclear power, and electric propulsion (ion or plasma engines). Nuclear fusion would be an improvement, but to really go fast, you want to use the Sun as a power source.

We have known for over a century that the Sun's mass bends starlight. You build a very powerful laser near the Sun, where energy is plentiful. You send the beam to a relay mirror farther out, then back towards the Sun's edge, where it acts as a giant lens to focus the beam. The beam powers a particle accelerator on your starship, which you point aft to speed up, and forward to slow down.

Since your energy comes from outside the ship, there is no limit on how much you can use, even more than antimatter could supply. Relativistic particles in your accelerator gain mass, something we do every day in accelerators on Earth. So your exhaust mass can be arbitrarily larger than how much you start with in your storage tank.

In theory, this method could reach near the speed of light. In practice it will be limited by how much energy your equipment can handle without melting.

> you might need to go 4000 ly+ for something even close to Mars conditions

What civilization does is take the natural environment, and make it more comfortable. Clothes and heating/AC keep me comfortable despite temperatures going from 22 to 95F locally. We have people living at the South Pole, the hottest deserts, and underwater in submarines. Notably, the Space Station has been occupied for 20 years now.

Before we solve interstellar travel, we will first have solved space colonies in our own solar system, both on the surface of bodies, and in orbit. The main difference between an orbiting space colony and an interstellar ship is an engine to move it around.

If you prefer the Earth environment where you live, fine, you can stay there. Some of us like an occasional change of scenery.

5.2 Names and Descriptions

Launchways - Power-Assisted Runways for Space Launch - Includes gas pipe, electromagnetic, and mechanical accelerators.
Orbital Power Relay Network - to relieve vehicles and systems from providing their own power sources. Includes beaming power up to orbit, down to Earth, and between locations in space.
Space Truck Stops - Food, fuel, and rest on the way to anywhere. LEO, HEO, MTO, Phobos, Mars.
Satellite Service Station - by analogy to automobile service station. Fuel, repairs.
Space Engineering Task Force - To assign space addresses (by analogy to IETF)
Outer System Mining - Determine delta-V, mass return ratios, and materials availability.
Sustainable Mars Colony - By that we mean it is affordable to sustain it. Thus the whole logistics and production chain has to be made affordable. Pushing everything from Earth is too expensive.
Ownership of Locations - Establishing ownership tenure as a spur to space development. This includes defined orbits, as well as traditional land-based parcels.
"Developing Space: the Solar System and Beyond"
"The Riches of the Universe are All Around Us"

6.0 - Random Items

6.1 - Misc. Space Notes

Ground Out Radiation Belts

Radiation Hazards

Elevated risk @ 20% per Sievert Background radiation = 2.4 mSv/yr = 0.165 Sv/lifetime Therefore = 3.3% mortality via background radiation.

Total release = 370 TBq = 3.7 x10^14 disintegrations/sec I-131 half life = 8 days. Thus 5.1 million x 10^14 disintigrations I-131 decay energy = 1 MeV = 1.6 x 10^-13 J = 18.9 x 10^20 disintigrations = 30 x 10^7 J = 3 millimole I-131 = 0.4 grams I-131 1 Sievert = 1 J/kg = noticable health effects, ~3 Sv for death. 55 kg/ person = 550,000 person-Sievert at 100% efficiency.

Fukushima province outside exclusion area = 2 microsievert per hour = 16 mSv per year

Bikini Atoll: 76 Megatons in two dozen bomb tests 1954-1958 IAEA report: http://www-ns.iaea.org/appraisals/bikini-atoll.as

Orbital Mining

I call it "mining the Debris Belt". Space debris is all aerospace-grade materials. In many cases dead satellites have working parts, except they ran out of fuel or one critical part failed. So it makes sense to salvage this stuff and reuse what we can, while also cleaning up the hazard.

The trick to doing this affordably is a large supply of propellant. This can come from the ground, if SpaceX's heavy lift rocket works, from asteroid mining, or by mining the upper atmosphere from orbit (scoop mining).

The reason you need a lot of propellant is the debris is scattered in random orbits, and it will take a lot to chase it down, even if your "garbage truck" flies an efficient route to pick up multiple items.

Space Elevators

Monday, May 21, 2012 at 12:02pm EDT Danielle Eber commented on Space.com.

Sub-orbital elevators (which don't do the whole job of ground to orbit) are quite possible with current materials, and also quite useful by reducing the job of a launch vehicle from 100% to something like 70% of orbital speed. They also would not look like the artist's concept. Natural and artificial debris require that you build it with multiple redundant cables with cross-strapping to distribute loads around a break. When a meteorite or dead satellite inevitably hits, you only need to replace a small section of one cable, and not the whole thing.

Ceres Catapult

Centrifugal catapult on Ceres can launch cargo on direct transfer orbit towards Mars. Carbonates and water on Ceres can build up Martian atmosphere.

Starship Cost

Fuel is the least expensive part of a rocket. Methane is $230/ton and oxygen is $85/ton. The SpaceX Raptor engine uses 3.8 oxygen to 1 methane, so the blended fuel cost is $115/ton. Conventional aerospace hardware (787-10 airplane) costs $2.5 million/ton.

The SpaceX Starship is less complicated than a passenger airplane, and is being built in a field, so lets assume it is ten times less per ton. $250,000/ton is still way way more than the fuel, even if you need 15 tons of fuel for every ton of rocket hardware.

That's why the goal of the Starship rocket is to get back both stages and throw none of the hardware away - it's the most expensive part.

In terms of fuel cost per payload, the two stages have a combined mass of 4400 tons of which 100 tons is payload and 286 is hardware, leaving fuel at 4014 tons. That costs $461,600 per launch, or $4.62/kg payload.

All the other costs: amortizing the rocket hardware, maintenance, launch pad operations, etc will bring up the total cost per flight to a likely $20M per launch, or $200/kg.

Space Tower

Rigid arm on top of tower, with jet engine driver at tip. Cable hanging from tip has payload attached at end.

Drake Equation

Wednesday, January 11, 2012 at 3:08pm EST Danielle Eber commented on Space.com.

We have finally nailed down one of the factors of the Drake Equation: f(p), or the number of planets per star. The next step will be to nail down the fraction which can potentially support life. For that we need to finish exploring our own Solar System, and get better data on the orbits and masses of extra-solar planets. This will keep us busy for a few decades. After that we use the Sun itself as a gravitational lens to get a close look at those planets. That requires getting out to 1000 AU (1000 times the Earth's distance from the Sun), where the starlight comes to a focus. Since the Sun is millions of time larger than the largest optical telescopes, the detail you can see is that much greater.

Development Sequence

Monday, January 30, 2012 at 6:00pm EST Danielle Eber commented on Space.com.

It would be wasteful and expensive to build a Moon Base before we have the proper infrastructure to support it. First learn to retrieve NEO materials robotically and extract useful products from them. Build up our capability to support ourselves starting in Earth orbit, then gradually working outward to the Moon. Otherwise it's like trying to cross a jungle on foot. Yes, it can be done, but it's a hell of a lot easier if you build a road first. Once we have worked up to having a manned base in Lunar orbit, then we can send the robots ahead to prepare the Moon Base, controlling them from orbit. It will end up costing a lot less in the long run if we learn to "live off the land" as we go.

Tuesday, May 8, 2012 at 3:27pm EDT Danielle Eber commented on Space.com.

"RC Robots First" is an excellent strategy if the robots are controlled by humans, and are used to prepare the way for human visits. For a simple example, setting a fuel processing plant on the Moon or Mars, operating by remote control, so that by the time the humans land, there is fuel to come back. That eliminates the need to bring the return fuel with you, and so lowers the entire cost. We can practice this kind of operation in a hostile location on Earth, such as the Atacama desert, which is cold and dry, so similar to Mars, and the operators in the USA.

Extent of Solar System

The Solar System is growing faster than we are exploring it. Make graph of most distant

aphelion & voyager distance vs. time

Monday, June 18, 2012 at 2:41pm EDT Danielle Eber commented on Space.com.

The Voyagers may be getting past the Solar wind, but they won't have "left the Solar System" until they get past all the Centaur and Kuiper Belt objects. The current record holder is a 100 km object that gets out as far as 2194 AU, or 18 times as far as Voyager 1: http://www.minorplanetcenter.net/db_search/show_object?object_id=2012+DR30&commit=Show

Asteroid Value

The value of a resource is net sales price after mining costs. Speculating on values based

on current prices is silly, since increased supply will affect price, and ignores mining cost

Wednesday, January 18, 2017 at 11:29am EST Danielle Eber commented on an article.

No it is not worth quadrillions. This asteroid, if pure iron, contains 7 million cubic km of it. The Earth's crust contains a hundred times as much, and would be far easier to mine for use on Earth. At present rates, the crustal iron would last until the Sun causes a runaway greenhouse effect, like Venus, and destroy all life here. For use *in space*, there are much closer metallic asteroids, and the quantities we need are tiny compared to what we use on Earth. There is a market there, but it's a small one. A resource is only worth what you can sell it for. We're talking a few $ billion.

Solar Lens

The Sun is a giant gravitational lens, but the image is projected on a sphere that is more than 550 times farther than the Earth is from it (550 AU). That's a huge area to cover.

In order to image more than one target, we can make use of distant solar system bodies from the Kuiper Belt and Scattered Disk. We have already discovered a couple dozen objects whose orbits reach into this region. There are likely hundreds more that we simply can't find yet because they are too dim. These bodies will likely have lots of ice, because pretty much everything that far out does.

So we mine them for propellant, and use nuclear-electric powered probes to move around and get views of different targets (stars and planets mostly). If they run low on propellant, they go back to the mining site and get more. Due to the vastness of the region beyond 550 AU, you will want dozens of mine sites and hundreds of probes to get good coverage.

The probes would be relatively simple. They would have a line of sensors on a long cable that rotates. That lets you scan a large area of a planet or whatever over time. They relay the data to the nearest mining site, which has a bigger and more powerful transmitter and antenna to get the data back to Earth.

We won't attempt this until space mining closer to Earth is well-developed, but once we do, we can get detailed images of anything of interest out there.

Launch Costs

Some people have grasped the implications ever since the mid-1970's. That's when Gerard O'Neill started studying space colonies, and the Department of Energy started funding work on Solar Power Satellites. Both of those assumed the Space Shuttle's cost would be as-advertised ($61.4M for 29.5 tons to LEO in current dollars), and that a follow-on heavy launcher would be even cheaper. The much reduced costs allowed thinking about big space projects.

As we all know, the lifetime cost of the program ended up being $1.64 billion/launch, so all the grandiose plans got shelved. Fast-forward to today, and the Falcon 9 has an advertised price of $62M vs an actual max payload of 15.6 tons (their price page says 22.8 tons, but that's for fully expended, which isn't comparable to the reusable price). That's close enough to the Shuttle's original goal to make somewhat large projects like Starlink feasible (3,120 tons for the approved 12,000 satellite constellation). The most that the Shuttle and expended rockets could manage was the 420 ton ISS.

When I worked at Boeing, we studied the expected effect on satellites if launch cost went down. Our reason is that when you design a new rocket, you assume a "traffic model" of how large the payloads would be, and how many of them there are to fly. You then optimize the rocket for minimum cost. The simple version is that the marginal cost of making the payload lighter should equal the marginal cost of launching, both in $/kg. That is complicated by things like some payloads being bulky but not that heavy and launches come in discrete increments of size and mass. But the general trend holds.

So as launch cost comes down, you spend less effort optimizing for weight and use cheaper but heavier components. The other thing that happens at various cost thresholds is new business models become feasible, or markets dramatically change in size. Thus at $60M/seat, you only get a literal handful of space tourists. At $100K/seat you might get thousands or millions.

One business model that isn't yet feasible is off-planet production. Ultimately 98-99% by mass of a space project could be built from materials and energy already in space. Of course, it will start much lower. The "mass return ratio" (kg processed/kg equipment) is typically in the hundreds or thousands. Think of how much dirt a backhoe moves in its life, or a chemical plant. So the overhead of launching the needed production equipment should not be a roadblock.

There are two implications for off-planet production. The first is that more massive projects become possible. The second is how much launch from Earth you need for a project can be reduced as much as 50-100:1. Outside of a small community of forward-thinkers, not many people have thought beyond mining water for propellants & life support, and bulk rocks & dirt for radiation/thermal/micrometeoroid shielding.

Lunar Sublimation Bricks

No need to use so much focus on sintering silicates. During other extractive processes, many substances are going to be sublimated directly from solids into gases in vacuum. Those gases can be vented through loose aggregates. When they are cooled by contact with those surfaces and solidify, the material will be cemented, and with no environmental processes to break them down over time. With the correct forms, you'd get masonry units, or any other monolithic shape.

There are a few minor challenges, such as that the aggregate substrates warm through use, and sublimation will slow. As sublimation progresses, pneumatic conductivity will decrease through the aggregate, creating objects with different properties.

It is possible to apply multiple sublimation layers to the same aggregate, creating materials with interesting and potentially useful properties. ie, batteries with low power density

Bootstrap Space Factory

The first lunar "product" will likely be bulk rock thrown into orbit. The energy to get off the Moon (1.5 MJ/kg) is much less than the to convert rock to useful product (typically 10-30 MJ/jg). Solar energy is available half the time on the surface, but all the time in suitable orbits. So your processing and fabrication will happen twice as fast in orbit.

You can use a rotary catapult (a motor and rotating arm) for low volumes, and a linear electromagnetic one for high volume.

There are 2 main types of lunar rock (maria and highlands) and 3 main types of asteroid rock (caronaceous, stony, and metallic). All five represent different "ores" with different composition. For a full industrial capacity, you will want to use all of them.

There are 24,000 known near-Earth asteroids, and the number is growing fast. Some of them are easier to reach than the Moon's surface. An electric tug can haul rock from them back to our space factory.

Combining carbon and iron alloy from two asteroid types we get a decent grade of steel. A starter set of metalworking machines can be fed with this steel to make parts for more machines. The second generation machines would be optimized for working with other materials and chemical processes. That way we can bootstrap industry from a small starting setup.

The SpaceX Starship will be able to deliver 100 tons at a time to the lunar region. So it won't take many flights to deliver the starter factory.

Space Tethers R&D

For space tethers, an example is a "Variable-G Research Facility". The ISS has given us plenty of exeperience with zero G effects on people and other things. But we have essentially zero data on gravity levels between 0 and 1. So we could use a VGRF to do the needed experiments. This would consist of a structural truss, which could be rotates at different RPMs, and movable experiment modules that can be placed at different radii as needed.

Once you have such a facility, you can dangle various kinds of tethers from the ends of the truss, experiment with reel mechanisms, current and plasma contact techniques, cable dynamics (a long cable flexes a lot), etc. This has merit from a research standpoint. Once you have done all you can at that level, then you can think about prototype free-flying tether systems for smaller applications, whatever those are. An example would be to chase and catch a piece of debris, lower it some number of km on a tether, and let it go. Perigee will be 7 times the length of the cable lower, for rapid debris de-orbit, while at the same time the chase vehicle gets kicked to a higher orbit for the next target.

Where you go from there is too hard to guess right now, but that should give an idea of the process to go from idea to real projects.

Space Debris

Space debris de-orbit gun (vertical launch air or water in path of stuff)

6.2 - Other Subjects

Decarbonize Shipping

Offshore wind turbines are a small but growing share of the wind energy business. So set up floating wind farms along shipping routes. Use them to charge up tugboats that consist of only batteries and motors. The containers ride in a separate craft that is pulled by the tugboats. When one tug runs low on power, it is replaced by a fresh one, and the depleted tug goes to the nearest wind farm to recharge. This avoids having to carry enough batteries for the whole trip, which would be heavy and expensive.

Alternately or in addition to the above, sail power was used for hundreds of years. If you redesign the ship to a catamaran hull, you can carry more sail area, and go faster than historical ships, and modern sail designs get better performance. You will still go slower than most cargo ships, but you avoid fuel and battery costs.

A final method is to develop kelp farming. Some seaweeds have a high lipid content and can be used as a biofuel. Some combination of these may work.

Freezing CO2 From the Atmosphere

Antarctica is already very cold. The atmosphere above ground level is even colder. There may be some combination of temperature and pressure where CO2 naturally freezes out, or is very close to that condition. The concept is then to cool the air as needed, freeze out the CO2, then transport and store it somewhere for the long term. The benefit relative to direct air capture in warmer regions is potentially low energy required to separate the CO2. The detriment is having to work in Antarctica. Whether this is cost-effective is unknown.

6.3 - Humor

Asset Droid Mining (a pun on asteroid mining).