A taste of things to come

Of all the institutions embedded in the culture of science writing, abstracts are actually a good idea. Abstracts give the reader a broad overview of the entire document – it’s structure and arguments. Essentially ‘warning traveller!’ A good abstract is like a tree in winter – there won’t be any surprises about the structure come spring, but you have to wait to see exactly how it will be filled out. This is not an abstract. The idea (right now) is to give some idea of what is to come and populate it with links and retroactive changes as this thing evolves.

In one sentence: This research presents the case for contactless spacecraft actuators utilizing eddy-current forces and an approach towards developing them that combines high-level modeling and empirical testing.

First, why the non-existence of eddy-current actuators is a problem worth solving and caring about and what science and engineering has already been thrown at it. This background will cover, in increasing specificity, actuators, spacecraft actuators, contactless spacecraft actuators as well as eddy-currents and earth-bound eddy-current actuators.

That will cover everything outside of this project. Then, an overview and justification of the specific angle of attack project takes – a general (gross) model combined with empirical testing.

From there, the nitty gritty.

Theory: different tactics for producing eddy-current forces in each of the three spatial degrees of freedom (push/pull, side-to-side, up/down), how an eddy-current actuator can be used as a sensor and actuator at the same time with no additional equipment. Algorithms for modeling a system that ties together Maxwell’s equations and rigid body dynamics.

Experiments: How strong is this actuator? How do we wrap a feedback loop around it? How much can the model predict? How to test a spacecraft actuator both on Earth and on the cheap?

And of course, where it can all go in the future (besides statistically likely scrapheap of history.)

For Science!

I’m going to try an experiment. The goal? A steady stream of posts that will be like dissertation Legos. Little discrete blocks of thought that can stand on their own, but snap together to form something far more awesome.

Clearly posts will always be far less polished than a complete dissertation. Many technical details will probably be glossed over. Breaking a dissertation into blog posts is pretty unorthodox and perhaps heretical.

But what is a dissertation? From Wikipedia: 

Dissertations normally report on a research project or study, or an extended analysis of a topic. The structure of the thesis or dissertation explains the purpose, the previous research literature which impinges on the topic of the study, the methods used and the findings of the project.

To that, add the idea a dissertation is where you show that you have expanded human knowledge just an iiiity bit. There is no reason why all of these objectives can’t be accomplished brick by brick.

I think there may be couple of big advantages too:

Thought-bricks build up a document that can quickly link to the origins of its ideas directly, rather than leaning on footnotes, references and appendices.

I’m a huge fan of iteration and bottom-up design, and thought-chunks allow me to put that bias in action. Most dissertations pop out of their author’s heads like Athena, fully formed and raring for battle. Although they can be edited in small ways, it’s much harder to fundamentally examine the organization of ideas in a longer document.

This format allows me to expose my nascent ideas to the world for inspection like Spartan babies. With your help and comments, not-so-gentle-reader, hopefully only the strong will survive.

Engineering vs. Economics (a response)

Last week, Russ Roberts (of ever excellent Econtalk) wrote a blog post about economics and engineering. I was surprised by how much I disagreed, though not in the way you might expect.  His point: looking at the economy isn’t anything like an engineering problem. The economy is complex and emergent while engineering problems can be modeled very accurately by simple equations:

“Running an economy is not an engineering problem. There are no simple equations that describe its motion that are akin to the engineering problem of space travel.”

 The gap between economics and engineering is not so large as Russ makes it out to be, but the similarity is in the opposite direction from the normal misconception: rather than economics being captured by simple equations like engineering, engineering is more complex and emergent than most people (even engineers!) acknowledge, like economics.

 Sure, engineering theory can be captured by simple equations (“it’s just physics!”) but then, so can economics. It’s just easier to get to ‘the emperor has no clothes’ point in economics. You can write down the rocket equation, F = MA, do some orbital dynamics and say, “yup, that’s how we get to the moon” but this is what the wiring of the Apollo computer looked like:



Just as in economics, the gap between neat equations and reality is huge. Non-optimal solutions are built on non-optimal solutions because technology has become so complex that you can’t just start from scratch. A huge amount of programming is done using Unix command line commands and the C programming language, both built in the 70’s and flash frozen into the system.

 “Ironically, in an unplanned economy, shopping is usually as straightforward and predicable as space travel. It becomes a engineering problem–what is the shortest route to the grocery–what is my optimal path through the store given my shopping list.”

 From my perspective, Russ’ wonder at how well the unplanned economy works, rings just as true for engineering.