The text below is based on an idea that’s been rattling around in the back of my head for years but it’s only recently that I’ve started putting words to it. I don’t really want to label it a short story because it doesn’t have a proper ending, but I don’t really know what else to label it as. While it is kind of satirical in nature and it does borrow some concepts and ideas from the bible, it’s not meant to be an insult against God or Christianity or even religion in general. Instead I like to think of it as a kind of story-esque thought experiment. This is how I think that a god with an engineering mindset would go about creating humans. Note that the god in this story is some sort of disembodied force that doesn’t have any concept of anatomy. Note also that he hasn’t created any other creatures, he’s starting with humans.
Build log day 1:
I’m hereby starting a new project. The aim is to build some kind of machine that is able to operate completely independently without any tether. This means an onboard power source as well as a fully embedded control system. The task for this machine is unimportant but for the sake of proof of concept, I have set the goal for the machine to “moving around without crashing into anything”. If I can succeed in that, I plan to make it interact with the environment in some way but I leave the specifics of that to the future. I thought about calling the project Dog, as in my own name backwards, but after thinking about it for a while I settled on Man as short for Mechatronical Android. I might come to change that in the future, but I will at least use it for this first prototype. I’ve never worked on a project like this before so I’m not going to do too much design or planning beforehand. Instead I will be figuring things out as I go along, and solving any problems as they turn up.
Build log day 2:
Today I started work on the main body of the machine. I built a frame out of one inch square steel tube. Since I haven’t selected any of the internal components I couldn’t build it to exact measurements so I made a basic box shape. I estimated that around 80 * 50* 20 cm should be enough to hold all the components that I need to fit inside. The frame is very sturdy and should easily withstand all the forces exerted on it. Unfortunately it became a bit heavier than I had hoped but I can deal with that later if it becomes a problem. The next step is to find some actuators to make it move.
Build log day 3:
After a lot of deliberation I have decided to use four limbs, one at each corner of the main body. By moving the two diagonally opposite limbs at the same time, it will get a dynamically stable and still reasonably fast gait. In case this control scheme should fail I can fall back on moving one limb at a time which gives it a gait that’s completely statically stable. After making that decision I spent the rest of the day researching actuators. At first I thought about using electrical servo motors but then I found a series of electrically controlled hydraulic actuators from a company called Muscle Inc. that I believe will work really well. I have ordered a set of those to try out.
Build log day 4:
I got the hydraulic actuators today. I used a hinge to attach a section of one inch square steel tube to the main body as a test limb then connected one of the actuators to it. I don’t have a separate hydraulic pump so I jury rigged the actuator to run off the pump on my hydraulic press then set up a simple Arduino interface for sending the control signals from my computer. At first it didn’t move but after fiddling around for a bit I got it to work; the limb moved just the way I wanted it to. Based on this successful test I have decided to use these actuators throughout the project. That means I can’t keep relying on my jury rig so I started looking for a hydraulic pump. It turns out that Muscle Inc. has a series of pumps called Heart that are suitable for their actuators. I have ordered one of those, as well as a selection of hydraulic actuators. Hopefully all of it should arrive tomorrow. While I’m waiting I will work on the remaining three limbs.
Build log day 5:
I finished construction of the remaining three limbs and attached the Muscle hydraulic actuators to them. The Heart hydraulic pump arrived in the afternoon. I installed it in the frame and connected it to the limb actuators. I still don’t have any of the peripherals for the pump so I connected it to my work bench power supply, and used an old plastic bottle as a hydraulic fluid reservoir so I could run a test. All the limbs could move but I ran into a problem. The two limbs on each side will collide with each other, severely limiting the range of motion. I was hoping to work on a power supply for the pump but I will have to solve that problem first.
Build log day 6:
I spent the morning working on the limb collision problem. My solution is to mount one set of limbs inside the other so that they don’t interfere. To do this I have extended part of the frame outwards to form a kind of tip then mounted two of the limbs here. Since the limbs are now different I have decided to call them limb pair A and limb pair B. Pair A is mounted on the original frame while pair B is the one mounted on the new extension. I spent the afternoon figuring out what power source to use for the hydraulic pump. The simplest solution would be to use a lithium-ion battery pack but I was considering to use a hydrogen fuel cell instead. The commercially available fuel cells I could find didn’t meet my requirements. I was about to start looking for a battery pack and charger when I found an interesting alternative. A company called Metabola produce a system that can convert a large variety of biomatter into energy. I have ordered one to try out, if it doesn’t work I will revert to using batteries.
Build log day 7:
I have been testing the new energy conversion system from Metabola. At first the output was really low and I could,t figure out why. No matter what type of biomatter I input, the output was only changing by a few percent. After reading through the documentation in detail I found the reason. The process needs to be aerated to function at full efficiency. I tried using a simple low voltage fan to achieve this which noticeably increased the output, but not enough. It simply couldn’t generate enough pressure. I thought about installing a compressor but I’m afraid it would make the system too complex. Instead I found a set of bellows from Swedish company Lungor AB that should work.
Build log day 8:
This morning I took delivery of the aeration bellows. They require a sort of pumping motion to work which I can generate by the use of two hydraulic actuators running in synch but with a 180 degree phase shift. I made a small hydraulic circuit with these two actuators connected directly to the pump. I then hooked them up to the bellows and ran an air tube from it to the energy converter. Since this setup is circular with the converter feeding the pump which runs the bellows that feed the converter, I installed a small three-way switch between the energy converter and the pump. That way I can run the pump on wall power for a few minutes until the converter gets up to speed. At that point I can switch it over and the whole system should run independently as long as I keep feeding biomatter into the converter. I did a quick test run on the bench to make sure everything works then installed the whole package into the machine frame. That’s it for today, tomorrow I will run some more tests on the system.
Build log day 9:
In the morning I ran a few basic tests on the system, everything seems good so far. After lunch I wrote a small test script that can log some key data about the system via the Arduino control interface. I then set the system to run continuously over night to test reliablity. That’s it for today, I will check the logs in the morning.
Build log day 10:
When I came down to the workshop this morning the air was filled with a really foul smell. The source turned out to be a small pile of brown paste under the still running machine. I quickly turned it off then set about cleaning up the mess. Once everything was clean I could focus on figuring out exactly what had happened. The explanation is that the energy conversion process produces a sort of semi-solid waste product that consists of biomatter that cannot be fully processed into energy. I had simply overlooked this fact when starting the test yesterday, so this waste product that had been excreted onto my workbench. To prevent this from happening in the future I attached a large diameter rubber hose to the waste output port then capped off the other end of the hose with a manual valve. Next time I am running a long term test I will have to remember to dump the waste at regular intervals. Having taken care of the mess, I could check the test logs. All the values fell inside the acceptable range and stayed stable throughout the test. Except for the waste product incident I call it a success.
1 Pingback