Introduction: DIY Factory

About: bicycles, gardening, and other important stuff

I've built 2 off-grid, digital, open-source factories out of 20' shipping containers, and I want to help others build their own.

Shop-In-A-Box #001 operates in Bungoma, Kenya; #002 lives at American Steel in Oakland, CA. #002 travels: it's been to Maker Faire (while under construction), The Gates Foundation Reinvent the Toilet Challenge, and Burning Man in the last year. Here's an article Gizmodo did on this project; we cut their logo out of 18-gauge mild steel while on the playa.

Each factory cost ~$30,000 in total, including power generation and tooling. This means that for around the price of a nicer new car, you can make many aspects of a car: sub-millimeter-precision cuts through 3/4" metal, welding steel up to 1" thick / aluminum up to 3/8", a wide variety of metalworking / woodworking tools, and a lot more.

Off-Grid: we stabilize grid power when it's available. If it's not, solar panels power prototyping needs (computer, lights, fans) and a generator powers actual production work (compressor, plasma, CNC, etc.)

Digital: the heart of Shop-In-A-Box is a 4' x 4' plasma CNC cutter. This lets us easily roll out design changes between #001 in Bungoma, Kenya and #002 in Oakland, California.

Open Source: the design of the factory is open source (CC-BY-SA, specifically), and we're utilizing open tools wherever practical. I want to get to where we have open-source versions of every single tool in the shop, but I'll need *a lot* of help from y'all to get there.

Factory: this works as a tinkerer's paradise, but it's also meant to cost-effectively produce meaningful quantities of real products. Our factory in Kenya can produce ~600 of our biochar kilns (cut out of 18ga mild steel) in a month. Our kilns in Kenya retail for $45; with ~$5 per kiln going towards recouping capital expenditures on the factory, we achieve payback with less than a year of at-capacity production. Our U.S. factory was used to complete most elements of 2 Kickstarter campaigns in the last year: Black Revolution, our carbon-negative soil, and growerbot, our social gardening assistant.


My friend Jason started re:char, a company about biochar, and he wanted a way to get his charcoal-making kilns to farmers in Western Kenya. He didn't want to buy a 10,000-unit run of one design and send a container full of them from China to Kenya, wasn't happy with the aesthetic quality / consistency of 'juakali' (side-of-the-road blacksmiths) in Kenya, and noticed that I had a shipping container in my yard while also spending lots of my time welding together chunks of metal into pedicabs. He asked me if we might be able to send a shipping container full of tools to Kenya and have it turn into a shop.

A hectic 2 weeks later, we unloaded a uhaul of tools into a container in Houston.

Flash forward 2 months, skim over many customs headaches, and Shop-In-A-Box #001 started production in Western Kenya.

More broadly, lean and localized production matters because too many factories are large, stinking, expensive heaps of energy and land usage in countries with lax labor laws and minimal environmental standards.

Most shipping containers see a sad life moving consumer goods from China to here and metal scrap on the return trip. 

By combining the 2, we can get a saner and I dare say sexier way of making things.

Step 1: Safety

This is not another laser-cut plywood project: you're dealing with heavy, industrial equipment here. In the course of developing these containers, I have personally:
  1. Hit myself in the head with a piece of telespar that I was using as an extended lever on a farm jack.
  2. Gotten side-swiped by a container truck while rushing to get to Shop-In-A-Box (from lack of driving safety, not shop practices).
  3. Cut a tendon, ligament(s?), and knuckle in my left hand through hasty use of a cordless circular saw.
  4. Severed a 220-volt line by dropping a piece of sheet metal off of the roof of a container.
  5. Snapped a support chain while craning a container onto a truck with improper chain positioning.
  6. Created a fake-murder-scene by setting the container on an unopened gallon of red paint.
  7. Fires: many causes, numerous. Worst/best was a kamikaze generator.
Working with industrial-scale digital fabrication is literally worldchanging: you have an unparalleled ability to shape the real world around you. Take your time, think things through, and try not to unintentionally reshape your body in the process.

Step 2: Planning

Shipping containers are not large; they'll impose a healthy discipline on you to design your space.

Interior dimensions of a TEU (Twenty-foot Equivalent Unit) container are ~1169 cubic feet, 18'8"L x 7'8"W x 7'9" H. 

Key questions to consider
  • Size of materials: you'll be processing: you're not going to be happy trying to move a 4' wide piece of metal into a container that's got exactly 4' of unimpeded space between the doors and your destination. We found this out by deploying roughly what you see in 'factory layout' below: the 'generator, welders' and 'weld table' were moved outside after half an hour of trying to work with this layout.
  • Ventilation: You will literally suffocate if you don't exchange the dirty air you're generating with clean air from outside. 
  • Lighting: Seeing while welding requires *a lot* of light. I often use 2 100-watt-equivalent CFLs mounted at opposite ends of the container and a third as a task light several feet from whatever I'm working on, and I could definitely use more.
  • People: it's crowded with 2 working inside a container that only contains a 4'x4' table, downright difficult to move let alone work with more. Pay attention to the anthropomorphic data in the picture below (from Jay Shafer's Small House Book)
I'm not going to touch on zoning issues: it's controversial + outdated for residential, and industrial's even worse. Know the possible consequences of violating zoning rules, and talk to other makers in your area to get a sense for how/if these rules are enforced.

Save yourself a lot of cost and time by modeling everything in CAD before cutting. It's a frustrating several days to learn the basics of Inventor (my preference), Sketchup (lets you place models in google earth satellite views), or something else, but this is time well spent. If it's 'measure twice, cut once,' it should be 'model thrice, cut once':)

Step 3: Getting a Container

For a shipping container in good shape, expect to pay between $0 and $4,000. Seriously: if you're patient, sometimes you can find difficult-to-move containers for free.

Find somebody local selling them on craigslist; conex is an industrial term that some sellers use in their listings but most sellers don't know to search on. I bought my first container (of 4) for $1,200 delivered in Austin, TX; I'm convinced I got this deal because the ad showed up in searches for 'conex' but not 'shipping container'.

If you're shipping the container overseas, just buy it from your shipping company: it has to be certified, or cargo ships won't ship it. I used Finn Container, but I would NOT recommend them: little to no support with customs.

For more detail, see my Instructable on getting a shipping container.

Step 4: Moving Containers

You've got 3 main options to move your container:
  1. Overhead industrial crane: I am spoiled in that American Steel has overhead cranes built right into the warehouse. So easy I might call it cheating.  
  2. Portable power movers: hydraulic crane trucks are my preferred option for moving containers where you don't have built-in infrastructure. They're usually less than $200 per hour, available most places in the world, and typically come with a crew to do the actual moving for you. Be sure to check if the company you use has a minimum number of hours.
  3. Manual labor: I used wooden rollers and a come-along on my container in Austin; never again. Blood, sweat, and tears by the gallon.

I will seriously drop whatever I'm doing and help anybody who wants to turn a container vertical, make the right-shaped base, and try to walk it like a Moai:

Step 5: Foundation

You need to set your container on something, and I've tried basic kinds of foundations:
  1. Pier and beam: the beams are essentially built into the floor of the container, so one can support the container by placing beams at the corners. You can see my shoddy attempt at this below; it's literally just cinder blocks at each corner, sinking into the ground due to the weight of the container. 
  2. Slab: much easier is to simply set the container on concrete slab, poured and leveled just as you would with any other building foundation. We did this in Kenya: easy, cheap, and thus far (1 year) very stable / level. I'm doing the  same thing at American Steel: we simply set the container on asphalt.
Aspirationally, I want to try the following:
  1. Proper piers, lofted container: dig 8' holes for the piers, as you would with piers for a pier-and-beam-house. extend piers 8' above ground. Rent a crane and place container on top of piers. Build spiral staircase to access your container in the sky. Optionally, fence in ground-floor workshop:)
  2. 20' base, 40' loft. My dream home + shop is a 20' container base with a 40' container centered above it. 2 containers = 480 square feet interior space, 3 decks, and 2 shaded areas:)

Step 6: Storage

You probably already know this, but let me stress that you'll need way more storage than you initially think.

Luckily, shipping containers are tall and durable storage containers are cheap. I created shelving by welding 3/4" tubing into a long 'L', covering the base with expanded metal, and welding this directly to the wall and ceiling of the container (sand through paint first, do NOT breathe this nasty likely lead-based stuff).

The containers are rubbermaid 18-gallon roughneck, measuring 23.9"L x 15.9"W x 16.5"H. The shelving measures ~25"L x 17.5"W (to wall of container) x 24"H per chamber, with 2 smaller shelves on the end. Total shelf length is ~18'8".

Hugely important is organizing all your stuff. I've labeled individual levels of my tool stand (discontinued floor model -> ~50% off at Home Depot), rubbermaid bins, and each part drawer of my small-parts bins; contents are then inventoried in a searchable spreadsheet.

Here's my spreadsheet of parts and locations, thus far.

Eventually, I will expand this to voice-recognizing: "shop, where are my 3/8" fender washers?" will trigger an LED blinking behind the right drawer. Maybe even 'shop, I just used 10 1-inch wood screws' integrated with inventory management and forecasting to ensure I've always got an average week worth of consumables on-hand:)

Step 7: Tooling

The heart of our shop is a 4'x4' CNC, currently configured with a plasma cutter capable of cutting through metal up to 3/4" thick with sub-.01" accuracy. . By deploying a general fabrication setup, we're able to 'ship right, ship once:' there should be few to no parts we need to send from the U.S. after a Shop-In-A-Box is deployed.

The CNC in Kenya is a Plasmacam that we got used with 2 plasma cutters for $6,000 ; I'm happy with it, and I've also been happy with a 4'x8' Torchmate I ran. Both are way too expensive new.

Shop #002 has a 4'x4' table built from 2 kits. The table itself is the iplasma 4'x4' and cost $2495 delivered with upgrades (water table, cable carrier, stainless tubes, and support for 620 oz-in motors). The controller and motors are CandCNC's Bladerunner with 4 620oz-in motors, Hypertherm torch interface, and Sheetcam + Mach3 for software; this cost $1871 delivered. We are running Hypertherm Powermax 45 plasma cutters; they were ~$1,700 each. 

On later shops, I intend to go full DIY on the CNC table and controller.

Our other main tools are:
  • Solar panels + batteries / inverters adequate to power the shop during prototyping / design (computers + 1-2 hand tools).
  • Generators adequate to power the shop during production (usage of welders, plasma CNC, etc.).
  • Transformers capable of scrubbing irregular grid power to a state where it is safe for use with shop-in-a-box.
  • 2 plasma torches, one for CNC use and another for hand operation. Each is capable of severance cuts up to 3/4″ and sustained cutting in any thickness metal from 1/2″ down to 22-gauge.
  • Full MIG, TIG, and oxyacetylene welding setups for joining a wide variety of metals.
  • Electronics prototyping, centered around through-hole components and Arduino microcontrollers.
  • A wide variety of smaller tools: hand, power, and pretty much everything else you’d expect in a well-outfitted garage.
  • DVR with 4 cameras, mounted to easily capture and share all details of project builds. (*not installed yet*)
  • All computers and software necessary to support the shop.

Here's an inventory of what we sent to Kenya for Shop-In-A-Box #001 (pdf). #002 has essentially the same capabilities, with the addition of a 3d printer and desktop cnc router. (I'm not fully satisfied with either of these, both in terms of their functionality and the value that they add for our Kenyan factory: what can we make and sell at a profit that we currently do not by adding these machines?)

The *right* way to do this, IMHO, is 2 CNCs with fullly interchangeable heads between them. 1'x1'x1 desktop version and 4'x8'x2' shipping container / garage edition. Switch your plasma torch, extruder, laser diode, dremel, pick-and-place etc. between them; eventually, have the table capable of changing its own heads:)

Step 8: Power

We treat power with a waterfall approach: we use the lowest-cost source available and fall back to more expensive sources as necessary.

Practically, this means we run on grid power when it's working. When it goes out, we switch over to our 10kW generator if doing production and our 60W solar panels if in design.

I'd consider power transformers / regulators necessary for operation anywhere outside of G8 countries and a good idea everywhere.

If you've got $60k+ burning a hole in your pocket and get enough sun, go with a 20kW+ solar array, 1,000kWh+ battery storage, and a 10kW+ inverter:)

Step 9: Comfort

If you're going to have a personal factory, you're going to need to make it enjoyable to work in.
  • Furniture: make your own, sized to your body and oriented to your working preference (sitting vs. standing, etc.)
  • Paint: all my containers are coated with white, high-reflectivity roof coating to keep them from getting too hot in the sun. Having just painted the inside of Shop-In-A-Box #002, I highly recommend it: it feels way bigger than it did with the gray original. Also, this paint is supposedly more resistant to mold than the original coating (we had some serious growth after leaving the container sealed with the lights on and water from the water table evaporating for several weeks)
  • Floor: standard containers have some wacky plywood treated to make it insect-, flame-, and rot-resistant. It's really nasty to get a splinter from; if you care, I recommend floor tiles or a garage paint (may or may not work on wood; lemme know your results if try it) 
  • Airflow: get big fans, and get more than one. Plasma cutting and most other processes that are any fun produce nasty fumes. Circulate clean air in from the outside and push the dirty out. This is as if not more relevant to safety than comfort. Here's a useful EPA document on air quality (pdf); go well above their recommended CFMs per square foot, and keep in mind that a fan's CFM rating is going to be far higher than the actual amount fresh air it delivers in real-world conditins
  • Roof: the ladder mounted on the interior of my door is the most useful thing I have ever made:)
  • Cord management: I have extension cords with outlets every several feet run inside along the ceiling's edge
  • Orientation: this matters for solar gain and prevailing wind if you're outside. Whether inside or outside, think about the positioning of your container's entrance relative to other tools and/or workspaces that you'll access. If you're going to repeatedly use a press brake 
  • Climate: toughness combined with attention to the above will be enough for most people. If you're getting hot in the container, get more fans: more airflow is going to improve your air quality, and the only downside is that you'll scatter papers. I don't like living in cold locations, so I haven't had to deal with a heating system. When I do, it'll most likely be a small, DIY wood stove surrounded by some rammed earth for thermal mass and located near the middle of the container.

Step 10: Security

You need to think about security of your container while it's being transported and when it's in use.

During transportation, you want everything secured to the actual structure of the container itself. I have dozens of tie-down straps that I run between the anchor points on the container and/or to press anything I don't want moving against the wall.  The CNC table has a piece of angle iron added between the legs and screwed directly into the container's floor.

Also, think about adding further lockdown to reduce theft. Nothing large was stolen during Shop-In-A-Box #001's time in customs, but pocketable things walked away. Someone actually stole the RAM out of one of the shop computers!

After deployment, a simple lock or 2 on the container door goes a long way. I've added the chain so that I can keep the container open a bit, allowing more air circulation while I'm gone for the night with the tradeoff of reduced security. We've also added a perimeter fence and guard (~$10 per day) to #001.

We've had one break-in to Shop-In-A-Box. It happened in Kenya; someone came with a gun (rare there) and threatened our guard. The robber got away with a smallish 12-volt battery from the generator and nothing else. Big, heavy tools are hard to steal on a motorcycle!

Step 11: Expansion

As your factory's production needs grow, it s quite likely you will need more than 160 square feet. 

To create more room for work and machinery, I've created nesting subunits. 7'x7'5' cubes made of 1.5" square tubing with 3 sides covered in fencing and 1 side having a roof  fit nicely inside the container for transport; you can pack their entire volume with other gear by placing the covered sides against the walls and floor. To deploy, I bolt the cube to the outside of the container with carriage or other difficult-to-turn bolts, heads facing out and 2 bolts inside. It won't resist a determined attacker, but it does keep people from walking away with my generator or oxyacetylene tanks:)

Later, I hope to add 'specialized' containers for higher production volume and specialization: a full-container plasma CNC next to a full-container powder-coating oven, etc.

Step 12: Future

For shipping containers and factories, we're just starting to scratch the surface of the awesome possibilities.

On containers:
Restaurants, residences, art: tons of projects across the board are staring to appreciate the potential of these funky boxes. I'm particularly interested in reuses that respect the original form, like the art piece below (Maker Faire San Mateo 2012, I believe by somebody from NIMBY). 

On factories, and production more broadly:
Thousands of units produced around the world are less fun and often less sustainable. Dieter Rams' idea of 'Less But Better' sounds like a great inspiration for making things custom and local. If they're better and we buy fewer, I think it's ok for things to occasionally be a bit more expensive.

Step 13: Join In!

Get on it: start building!

The more tiny factories we get, the better we'll get at making them. I want this to grow into an 'API for hardware', where once one person has ported an Instructable or other project into our standardized tooling setup it becomes way quicker for everyone else with the same setup to complete the project. Cue utopian visions of decentralized production creating robust local economies with sustainability innovations shared around the world, hardware like software:)

If you'd like to see one of my factories and can get to Bungoma, Kenya or Oakland, California, send me a private message.

Please comment below with any questions, advice, etc.

Thanks for reading; hopefully, this Instructable will help you make better mistakes than I did when you build your own factory.

Also, 2 pictures of my dog Puck in Shop-In-A-Box #002 pre-conversion. Because he helped:)

Instructables Design Competition

Participated in the
Instructables Design Competition