Monday, September 15, 2008

Distributed Manufacturing and Disaster Revovery

Another weather update today: Hurricane heading for the Gulf Coast. It's the third time this month we have heard this announced. In the news report that triggered this writing, we hear that the contractors working on rebuilding from Hurricane Gustav are being asked to ramp up their 30 day contracts to repair damage so that the work can be done in a week or so. As this is being prepared for posting, we are hearing of the aftermath of Hurricane Ike and the destruction of Galvaston.

Three years ago, New Orleans was hit by hurricanes Katrina and Rita. The population of the Gulf Coast was scattered all over the country because the government was not ready to quickly house the people and rebuild the area's housing stock. In the years since, we have heard of the problems with 'FEMA Trailers' and the environmental hazards the represent. Once people move out of these trailers, where are the trailers to go? Stockpile them for the next emergency?

In the days after Katrina, I remember hearing that there was a sentiment that "Americans don't live in tents". Our people should not have cloth houses for even a while (unless there is hunting involved).

So we have a problem and an opportunity.

The problem is that we have large populations around the world that are located in high density areas, and these areas are subject to a variety of natural disasters. We have heard this year of hurricanes, earthquakes, floods, fires and tornadoes to name a few. Whenever these areas are wiped out, housing needs to be replaced, quickly and safely.

The opportunity lies in how the housing is replaced. What if we had a system to rapidly manufacture housing that is safe, sturdy, attractive and can be used for permanent or temporary use? This housing would not be built using traditional methods developed a hundred years or more ago, it would be an opportunity to reinvent housing design, construction and use.
Distributed manufacturing of architecture
With computer controlled tools like those found in the Fab Lab system and manufactured by
ShopBot, we could create a network, locally, nationally and globally to provide good, safe, rapidly assembled housing. There are hundreds or thousands of tools around the United States and world that could be used on relatively short notice.

The parts of the houses could be shipped to a distribution point, then trucked or trained in to the disaster zone. Once on location, the parts, which would have been cut to consistent specifications on similar materials. Obviously, some allowances would have to be made in the design to ensure that the parts would assemble properly. Assembly could occur with relatively simple tools like rubber mallets. The site would need to be prepped, a cement slab or some other foundation.

Each set of parts from a distributed manufacture would have markings that would help with quality control. It would be possible to check the numbers carved into every part and tell which machine it came from, and from there tell which material was used, and other data that the tool owner/operator would keep track of. This would help in the design process by allowing the testing of materials and techniques.

As the crisis is resolved, people in the emergency zone could continue to live in the housing, or they could turn it back for other housing. Since it is possible to build the house entirely with mortise and tenon, pressure fit, tab and slot techniques, the housing could be assembled with little or no hardware fasteners. If there is a surplus of these distributed manufactured housing units, they could be sold, moved or even cut up and chipped.
Will it blend? Yes it will.  But can we Watch it Shred?  Yeah, that too.

In preparation for the next emergency, a number of these housing units could be stored warehouse style, maybe in shipping containers that could be put on trucks or trains. When the next hurricane/tornado/flood/fire/earthquake comes, pop the houses into the transportation system and get people into safe housing quickly. These stockpiled kits would be from manufacturers who had been properly proven as having good quality control, since these would be the first to hit the disaster zone. In days, or maybe even hours, it would be possible to get roofs over the heads of the people who recently lost their houses to mother nature.

This would change the way we make houses. Instead of just making the housing faster, we would be reinventing the system of housing design, manufacture and construction. It would encourage collaboration and open source design. It would release the manufacture of the house from the local area, and support open source manufacturing. It would create a viable solution to the rapid and safe rebuilding of housing in areas where there are vulnerable displaced populations.

Larry Sass of MIT has been working on developing a system very similar to this. His focus is on the
building itself. His designs and those like his could form the base of the system. But between where we are now and having trailers loaded and ready to ship to the disaster zone, we would need to have several intermediate projects of lower stakes to prove out the system.

The test part is an important step in the design process of the project..  A whole set of test parts with calibration fitting points could be made on one machine that has a known calibration and a history of accurately cutting the project parts.  These tester parts could then be shipped out to the project participants who would then make sure their machine can make parts that fit into the tester's calibration areas.  It would have some mortise and tenons and holes and rectangles, and probably some more than that.  The part would likely be no more than an 8th if a sheet of plywood.  When the machine user can make similar parts that are compatible with the tester, then the machine is calibrated to the international standard.  The machine user may just want to know the machine is set right.

At that point,  they can make an example duplicate of the tester and ship it back to a quality control group.  If the part passes a test, then the machine is official.  The official part would be optional unless the machine would be used on long distance group projects where intermachine compatibility is essential.  The certification process could be done by shopbot, or some other group of trusted individuals.

As far as proving the system, it would be absolutely essential to test it out on some low stakes projects first.  This would allow the manufacturing team to go through the design process as many times as needed to get the system working accurately and effectively. If the project had a number of intermediate projects and goals to build up to the biggie of fabbing houses that would sit in shipping containers waiting for a disaster.  When you get to that point, everybody involved has to completely trust the system with good data behind its' testing.

Intermediate level projects could be the manufacture of something pretty simple but at least a bit complex like rubber band cars or a toy paddle boat or a simple chair/table/organizer/other piece of furniture.

As the project progresses, the manufactured projects get more complex or made of larger runs of parts.  The rules governing the design get more precise, so then you can get into items that can meet national or international standards for manufactured goods.  

Ultimately the project gets to make structures, maybe starting with a dog house design or storage shed or car port.  When that system is working well, then the project tries out some even more complex designs

Shopbot Tools -
Blend Tech -

Labels: , , , , , , , , ,