Tuesday, June 9, 2015

A super short introduction to using GitHub for hosting a non-profit website.

GitHub is a for-profit firm that effectively has created a social network of open-source hackers.  It is now a "place" to go for hackers.  Fundamentally it is a repository for tracking, versioning, and retrieving files---but EFFECTIVELY it is a social work place for files.

By files, we mainly mean computer programs, which really require versioning.  But you can put anything there.  

GitHub makes it completely free as long as you are public/open source in your files.  You have to pay to have private files.  GitHub graciously gives 501c3 charities $25/month of services for free.  You can do this if we need a private repo.  For example, I am on a board where we keep private files, and also have a public face.

Additionally, GitHub implements "GitHub pages", which is a lot like WordPress or BlogSpot.  The main difference is that content is mostly not created with a "What you see is what you get (WYSIWYG)" editor. In general, you create content using a very, very simple version of HTML called "markdown".  

Here is a site that is implemented with GitHub pages: http://presidentialinnovation.org. The "repo" for it is here: https://github.com/presidential-innovation-foundation/presidential-innovation-foundation.github.io, where you can partially see how the site is constructed if you study.  Anyone can make a "pull request" which will be taken as a suggestion by the owning organization.

I think it looks rather nice.  To make it look nice, you have to do some extra work, but it is just "plain old HTML and CSS".  This is more or less easy.  By using GitHub pages, you are sort of using the "plainest" and simplest hosting technology that we can.

When you start with github pages, you get a URL like http://github.io/yournewsite.  However, you can point any domain you own at it.  That is how "presidentialinnovation.org" gets pointed there.

One of the great things about GitHub is that you can create an "Organization" which can then own "teams" and "repos", which are ways to organize projects and subproject.  This means that you can control who is empowered to post things on the site.

Someone will NOT be able to post things on the site unless they either:

*) Get a free github account and you add them to the "team", or,
*) They submit a "pull request", and someone on the team "merges" it.

This is how open source is organized.  ANYONE could make a pull request (that is, "request us to pull their change into they site"), but only YOU can merge it.  It is not uncommon for people to make pull requests to fix typos, or to augment information, for example.

I believe for many charities, clubs, and organizations, hosting a web site with GitHub Pages should be a defacto starting point.

Monday, June 8, 2015

A Cold Calculus Conversation with a Neonatologist

Today, following the surface-level success of the ATX Hackathon, I spoke for an hour with a person who had a lot of experience in Africa and Bangladesh with premature babies.

She informed me of a number of difficulties with the Premature Baby Warmer project, although she confirmed some thoughts that preserve a slight hope for the usefulness of the project:

  • Warmers are certainly needed.
  • A system that could use cell-phone batteries successfully is potentially valuable.
  • In most of the developing world, the Ministry of Health provides most of the supplies to clinics. This is a top-down model, and the idea of publishing an open-source solution is at odds with this, being a bottom-up approach.
  • It might be true that 10% of the babies that need it could be helped by people who have the kind of internet access that would allow them to use internet instructions to build a warmer.
  • Kangaroo Care is effective in approximately the same conditions that we are trying to address, so it may be that only a fraction of births, when the mother if disabled, benefit from this idea anyway.
In other words: if we manage to build an easy to construct $25 warmer than is 100% safe and effective, the number of babies we can hope to save with this approach is just a fraction of those that might need it.

Of course, perhaps more than 100,000 low birth-weight babies die every year.  Perhaps 10% of those will be close to someone with a good internet connection and the skills to build the Arduino-based warmer. Perhaps 10% of those will have a mother that, at least temporarily, needs a warmer.

One in a 100 of 100,000 is still 1,000.  I'm willing to put more time into a project that might save 1,000 babies a year.

Furthermore, it is possible that the know-how from the creation of this will help someone smarter build something better---or build something completely unrelated which is valuable.

So, without believing that I am very close to producing much good in this world, I think this project is worth some more of our time.

Contact me if you want to volunteer---there is no lack of work to be done, with almost any skill.

Report: ATX Hack for Change 2015 Preemie Baby Warmer

This weekend I led a team at the ATX Hack for Change Hackathon improving the Preemie Baby Warmer project of the Austin chapter of Engineers Without Borders.

Each year more than 100,000 babies born prematurely without access to Western medicine die needlessly. There are at least two attempts to address this problem.  The Embrace warmer is an elegant solution available in theory for $25.  However, it is made by a for-profit company, and its design is not open-source, and depends on difficult-to-manufacture phase change material.  I would prefer to empower local people to construct their own using inexpensive and readily available material.

Rice University's Oshman Engineering Design Kitchen has created a different solution: a $250 box-style isolette.  However, this is clearly an "in clinic" solution.

Engineering seeks practical solutions.  There is usually a design spectrum, and it is better to have a richly populated design spectrum.  We are producing a different point in the design space than either of these solutions.

I had previously constructed a prototype that I exhibited at the Austin Mini Maker Faire.  My friend Chip Rosenthal came by our table and invited me to the ATX Hack for Change Hackathon, specifically because I was working on a hardware project, something that would diversify the ATX hackathon. As one experienced hacker said, "I've never seen a sewing machine at a Hackathon before."

I was very lucky to get two brilliant young people on my team, and the help of an small team of IBM engineers as well. This is Josh Benson, who is a very clearly thinking and helped me with the electronics and programming, as well as just thinking about what to do.

 This is Cameron Lagrone, who was our MVP. She designed and constructed the "swaddle" completely from scratch in a single Saturday, and frankly it is more elegant and beautiful than anything I could have done.  She combined the sewing chops with the ability to think of innovative simplifications.
 Here is a picture of the "swaddle" that she produced, with our "Test Preemie" inside it.

It is not obvious from the photo, but inside the swaddle is a pocket holding the electric heating cloth controlled by the Arduino.  The tremendous beauty of this design is:

  • It can be made from two t-shirts and a shoestring an an hour or two of sewing.  That means it can be made so easily you DON'T have to make it before a child is born prematurely---which is of course unpredictable.  If you have the open-source instructions and pattern, you can make it before the mother needs to sleep from the delivery! And everyone has some cotton cloth and some string.
  • It needs no fasteners.  You just "swaddle" the baby almost as you would with a blanket. It imposes minimal cultural strangeness, as most of the world swaddles newborns with cloth in this way.
  • It is a separate module from the electronics.  You can use it just as a "blanky" without the electronics.
  • The "hoodie", when a fairly stiff material such as felt is used as the middle layer, protects the infants airway, while still insulating the head, probably better than a simple blanket could do.
The swaddle allows us to perform experiments we could not meaningful perform before.

The first observation that we could make is that, as I had previously conjectured, the Test Preemie that I created is a poor test object at present because it does not conduct heat significantly.

We fixed this with a quick trip to the pharmacy to get an gel-based hot/cold compress that weighs about two pounds (about the size of an underweight baby.)

A basic premise of this approach is that a large percentage of people have enough electrical power to charge cell phones, but don't have grid power.  However, the Warmer will only be practical if it can maintain body temperature for the infant with available battery power. This will require more research than we could do in the hackathon, but we could now move forward.

Oliver D. Rodriguez and the IBM team came to our aid here. They did something I have never done---they wrote an outside reader (in Python) which read from the output from the serial port written by the Arduino.  This allowed us to make a running data log of the temperature. (That's Oliver on the left most, with Kay Freund seated, who also helped.)


Using two 6 Volt lantern batteries, we were able to raise the gel pack from room temperature (25C) to skin temperature (35C) in 37 minutes inside the swaddle.  This was a successful test, although the actual use case is much easier---an infant should never have a room temperature body!  So I deemed this an encouraging result.

If you have ever been to a hackathon, you know how exhausting they can be.  I went home at 10:00 hoping to perform a far better "maintain the body heat" test---and collapsed.

In the morning, we finished our presentation, and mostly socialized, until the presentation.

We made what I consider a lot of progress:
  • We have a replicable set of instructions for making a practical, cheap swaddle.
  • We tested our ability to heat with encouraging (though incomplete) results.
  • We also added a sonic alarm for when the baby is too cold or too hot---which is a big improvement in safety and robustness.
I am pleased that the Austin hackathon is not competitive as some are.  Nonetheless, the judges gave out a few superlatives.  We won "Most Hacked Forward".

People seemed to generally really admire our project.  It felt great.

I now feel two things very strongly:
  1. It is my duty to document this as well as possible.  If others cannot use it, all of the energy and love is meaningless.  It is fine with me if people use it as an example to improve upon, rather than a tool for saving lives.
  2. In fact this approach is very far from changing the world, an a lot of work remains to be done.  I would estimate about 10 times the total work that has been done so far.

Thursday, June 4, 2015

The Agile Maker's Dilemma: How Shabby is the Right Amount?

I am currently trying to invent a new kind of linear actuator to enable producing the "Gluss" robots: omni-triangulated robots built only out of members that change their length.

I am dedicated to applying Agile software methodology to my hardware making.

I just spent a precious three hours trying to build a spool of the proper size to hold a homemade magnetic coil. Was this the right thing to do? (I'm writing this essay as the epoxy dries.)

The fact is physical making brings to the fore decisions that are not as difficult in software development.  Agile methods have been described as "The Scientific Method applied to software development".  That is, the fundamental thing Agile methods try to do is to learn things as quickly as possible.  Generally, this is "learn what makes the customer happy."

Kent Beck created Test-Driven Development as a way to be Agile without actually being shabby.  That is, in TDD, you are always in the control, and you always have a set of a running tests that define the correctness and quality of your system.

In a way, TDD allows you to always be "correct", but to always be developing on the dimension of completeness.  You start with a correct but very, very incomplete system.  You always stay in control by staying in control of your tests, so you're software is almost never "wrong", but it is constantly getting more "complete".  In other words, quality is always improving, and all errors are in the form of things not yet done, not mistakes.

This is a very successful strategy, in part because software, whatever other problems it may have, is extremely repeatable.  The same program always does the same thing with the same input.  Software doesn't really wear out or rot, or stick, or rust, or get gummed up, or overheat.

But physically devices, especially my home-frabricated ones, do.  In TDD software, you have a smooth, monotone improvement, with small refactoring, but there are more variables and more compromises in physical making.

If I wind my coil too broadly as I did recently, I have no recourse but to rewind it.  If my piston doesn't fit properly in my cylinder, I may be able to adjust it, but I may have to make a new piston or a new cylinder.  Either a bad fit or a bad coil is likely to prevent my fundamental test from being successful: the actuator may not move.

I do not mean that you should not try to make physical things via a series of small tests. But I believe it is harder to do this when physically making than making software.

The worlds really needs a book entitled "Test Driven Development for Physical Makers: How to Choose When to Invest Where for Maximum Learning."