day #2 November 4
11/03/06
âWell, you build stuff to figure out how it works, right? So that shouldnât be too hard. Just think of something that you donât know how it works. It doesnât have to be a device. Could be anything.â
That simple advice was all I needed to shake me out of my rut. Of course. âThe one thing thatâs always escaped me is biology. Biological organisms are built of entirely different stuff than electronics. Theyâre too... messy. Complex.â
âSo youâd have two problems: one, youâd have to reduce the complexity in the design in order to apprehend their function. And two, youâll have to figure out a way to increase the complexity of what you build in order to properly replicate said function. Thereâs a lot of research into the latter in software. How much experience do you have programming?â
âNot a whole lot, but Iâm comfortable enough with it.â
âSo you might want to look into genetic algorithms and fuzzy logic. And of course thereâs a whole bunch of study in artificial neural networks that will definitely come in handy for you. I think thereâs even been some work in hardware-based nets thatâd be right up your alley. I donât know how well they compare to software-based ones, though.â
âGreat. Iâve got something, then. This is definitely on an entirely different level than anything else Iâve worked on before, but thatâs really what I wanted, isnât it? And whatâs the worst that can happen if I fail? Iâm not doing this for anybody else, so I just start again.â
The rest of the time there we exchanged very few words. I jotted down notes and tried to work out how I would start designing my robot. Or android, I suppose, as it seemed only natural to create something in my own image. Joe, for his part, returned to his own research. I was a bit worried that he might be heading down a parallel path to his friendsâ, but I didnât say anything. He seemed to have a good head on his shoulders, and Iâm sure he could take care of himself.
The following day I spent at the library. I xeroxed as many anatomical diagrams as I could find, and gathered as many biology books as I could carry. Each and every organ in the body had to be transparent to me. And I knew that even though the brain would probably have the most research associated with it applicable to my own enterprise, it would certainly prove the most troublesome. So I tackled the other organs first.
But on closer examination, I decided I shouldnât start with organs themselves. I needed a more general overview. The body is divided into 12 systems. The skeletal system, digestive, muscular, lymphatic, endocrine, nervous, cardiovascular, reproductive, urinary, immune, respiratory, and integumentary systems. Other attempts at robots have ignored many or most of those, but I didnât want to take the easy way out and doing so would be a disservice to me.
The skeletal system would be the easiest. It mainly provides structure and overall form, and I could fashion it out of a lightweight metal or even a sturdy plastic. The digestive system would be quite a bit more troublesome. Did I really want this thing to have to eat? The main function of this system is to provide energy (and nutrients) to the organism. And to dispose of the byproducts. I figured as long as I closely mimicked the systems and remained faithful in principle, I wouldnât necessarily need to reproduce them exactly. After all, the idea wasnât to build an actual organic (in the chemical sense) creature. So as it wasnât going to have the same needs as a human, I would have to make alterations to the design accommodating these differences. Energyâs easy. Naturally, it was going to run on electricity. Any nutrients it might need Iâd need to leave to later on in the design process.
The muscular system would be one of the easiest, at least in terms of apprehending its function. Movement. That said, I would have a couple options open to me: the most basic would be simple motors. But motors are very different from the muscles in animals. Pistons or even titanium nickel alloys that constrict when supplied with an electric current would be better. Or maybe some other solution would emerge.
Next up I looked into the endocrine system. This is basically a messaging system, like the nervous system, except chemical. Several glands produce hormones which affect the behavior of cells in certain organs of the body. I imagined I might be able to combine this with my nervous system; it would simply involve a different class of messages.
The lymphatic system seemed mostly useless for my purposes, but I decided to study it with equal vigor. Even if I didnât end up creating an analogue in my robot, I wanted to understand it thoroughly to know exactly why I didnât need it. In function, itâs similar to the circulatory system except it deals with other fluids (lymph). It removes excess fluids from bodily tissues, transports fat to the circulatory system, and produces immune cells.
Which brought me to the immune system. I certainly wouldnât need to worry about pathogens, but I might modify this system to monitor the robot itself and possibly correct any problems that might occur autonomously.
Skipping ahead to another potentially non-included system I looked into the reproductive system. Now, organisms in nature reproduce either asexually or sexually, but as Iâm using humans as a template asexual reproduction seemed inappropriate. But as I was certainly not planning on creating a pair, a reproductive system would be pointless. Again, though, I didnât stop my research simply because it didnât apply to my problem. Like in art, where itâs important to understand the rules before one can (or should) break them, I had to make sure I fully understood a system before I could safely decide to exclude it.
And then the respiratory and urinary systems. I really didnât like where this was going, but neither seemed particularly useful for my purposes. The respiratory system is strongly tied to organic biology. Since my robot would almost certainly have no aerobic activity, requiring no oxygen for metabolism, a traditional respiratory system would be useless. However, I hadnât ruled out gasoline power so there might be need for an exhaust, in which case some sort of respiratory system would be useful. I would also need some sort of air intake / out-take if I wanted to mimic human speech and vocalization. The only use I could think of for a urinary system would be to rid the robot of any sort of soluble deposit that might form on its mechanisms, but at this point in planning it seemed unlikely that it would require one.
With only one organ system to go, and one of the most important â the nervous system â the librarian came by and told me it was time to pack up. Iâd have to leave the nervous system for tomorrow, but I had plenty to keep me occupied for the time being and I took out as many books on the systems I had so far.
Iâd spent way too much time today poring over anatomical textbooks so after I got home and dropped off the books Iâd borrowed from the library, I decided to go for a walk & try to clear my head.
word count: 1284 words
ugh. I didn't get nearly enough written when I got home from work tonight. Even more catching up to do tomorrow.