MJ Dogfood Session

Preparation:

Hi, my name is Marcin, founder of Open Source Ecology. I am working on the Global Village Construction Set (50 gvcs slide) - the set of the 50 Industrial Machines that it takes to build a small civilization with modern comforts. This is my second year of the Shuttleworth Fellowship. My best title would be something machine designer and Open Source Industrialist .

My goal is to create the open source economy - an economy based on open collaboration. While most people here operate in software and intangible goods, my work revolves around the 80% chunk of the economy that is the physical infrastructures of humanity - or material production - expressed through open hardware.

Why am I doing this?

O was born in Poland. When I was 7 years old – tanks rolled down my streets – and unfortunately, it wasn't a parade. I grew up behind the iron curtain in a state of martial law and material scarcity. My family and I waited in long lines for butter and meat. Then we fled to America, I got into Princeton and then finished a Ph.D. in plasma physics in 2003, and discovered- that I was useless. Also, I never stopped thinking about the terrible things that happen when resources are scarce and people fight over opportunity.

I said goodbye to my theoretical chalkboard and bought a farm in Missouri. I bought a tractor, then it broke - I paid to get it repaired, then it broke again - and pretty soon, I was broke - too.

That's when I turned to our roots of crowd funding. I blogged about all that I did, got subscribers and donors, and currently have about 500 True Fans supporting the project at $10 per month.

I built the first machine in 2008, and operated on a $1000/month budget until I was put on the world stage with TED in 2011. Since then, our budget has grown about 500% per year.

The first ever replication occurred in 2011 by a guy who quit his job as a programmer and built our brick press. In 2012, we have had over a dozen replications in 5 countries around the world.

2012 was a time of great growing pains, as I recognized the difference between vision and execution. Based on our meteoric rise, I proposed to Shuttleworth Foundation that I would finish the entire GVCS by the end of that year. They bought it hook line and sinker - not - and this year, I got a good spankin' as - Hail the King - is taking measures to make me more accountable to my promises.

Last year, we did - however, reach a major milestone. Our platform is Absolutely Efficient Production - Enabled by Open Source. Along these lines - we have achieved radical production efficiency of our machines. We build our automated, 1 ton, open source compressed earth brick press - in a single day.

Our goal for this year is to attain the same rate of production efficiency - but now on the design front. We would like to show to the world that it is possible to design a complete machine - in a single day. What we are talking about is disrupting the existing paradigm of proprietary research and development to the true promise of open collaboration - by creating a replicable, scalable paradigm of crowd development based on modular design. The concept is to break down a product into the smallest modules possible - to define the interface beween these modules - and to provide sufficient background information on the problem statement at hand - and to lower the barrier to collaboration to zero - by using tools such as Google Sketchup for 3D design, Wikis, and Google Docs.

How is this done? Via design sprints.

Our current idea is to scale this method to remote collaboration sessions with 50 people online. Once we reach 50 or so people, that goes beyond what one team can handle, so we will deploy other team leaders to run parallel design sessions - such as OSE Europe or other chapters.

We have shown that we have built machines in a single day. This year we aim to demonstrate that we can design machines - in a single day. And for 2014, our goal is to show that we can deploy an enterprise manufacturing machines or providing other services - in a single day. For 2015, our goal is to show that we can build out the 50 machines - in a single day.

Our general plan is to stabilize our organization this year - and we are currently developing our Design Sprints. We have just recruited 4 full time people - Operations Manager, Product Lead, Technical Community Manager, and Documentation Manager. In 2014-15 - we aim to engage an Apollo Project for the GVCS - to finish the entire GVCS by the end of 2015. Briefly on the plans after that - our goal is to shift to education as our core product from 2016 on. We will select 12 OSE Fellows for a 2 year immersion course on OSE from 2016-17 - culminating in open franchising to 12 locations globally. The last 6 months of this program will involve land acquisition and replication of a complete facility. We will replicate this in 18-19, to establish 144 of these facilities worldwide. That is Dunbar's number. My goal is that each of these facilities nets $1M per year from education and product sales, so that we establish a $100M/year funding mechanism for Open Source Ecology.

This may sound incredible, but it is what I live and breathe. Now I'd like to show you a glimpse of how Module Based Design and Rapid Prototyping work. We start with Sketchup to generate 3D files - [1]. We use 3D prints to validate designs and check for range of motion, such as for backhoe extension and curl. We convert the 3D to 2D for cutting on an automated CNC torch table. We model these cuts with cutouts from paper to make sure our conversion into 2D is correct - as cutting paper is cheaper than cutting 1/2" steel when prototyping.

We follow module based design so that parallel teams can build out parts. In this exercise, I will let you experience this from a 3D model that has been printed. There are 5 modules: power unit, wheels, loader, driver cab. We use 4 types of materials only: tubing, shaft, pivot plate, and connector plate. Now go ahead - break up into 8 groups of 3 and in 5 minutes we will build the entire tractor.

Here is the breakdown:

Dogfood Challenge

So now that we shot through the design exercise - I have a request on help on our main milestone for this year:

Achieving 1-day design times for machines. In particular - what is required for us to achieve 50 person design sprints, and to scale it to about 250 people in a day, via remote collaboration. What are the ingredients to make that happen? And, this leads to my second question how do we incorporate OSE Chapters to make this happen? Right now, we're seeing a number of chapters sprout up worldwide, and I get requests from here and there - India, Indonesia - to which I reply - we are developing our Chapters Policy this year. I would like to hear your suggestions on:

1. How do we attain the 250 person design sprints?
2. What is reqiured in terms of resources?
3. What is required in terms of management of these events?
4. Who are best role models for doing design sprints? Are there professionals that we can hire?
5. What should our Technical Community Manager be doing

Please use this google doc -