Test Department Onboarding

Auditory Log Print

Filter by Speaker<Keith_TestDepartment>;Flag<Notes>

Sort by Date

font:TimesNewRoman,fontsize:12,justified:justify

1// This is the breakroom

2// This is the life test laboratory. Each of these chambers is set up to run anything regarding the long-term performance of a product in certain applications. Typically, these tests are developed alongside the qualification campaign to cover extreme durations or cycles. The models here are usually identical to acceptance units, but sometimes they are manufactured with extra nerves so we have data related to functions that deteriorate too slowly to identify in a standard campaign. This one, for example, has receptors in all of its organs so we can estimate effective field time- even if we don’t run it until failure. Seeing how fast the kidneys deteriorate, for example, gives us an improved long-term effectiveness projection. You can imagine that the life tests are extremely expensive, so typically the only customers who ask for one are applying the product over a long period. This one is intended for terraforming, so the idea is that it will be released into unlivable atmospheres and its ecological impact will slowly make the planet habitable for humans. We simulate a day/night cycle and pump in ammonia to replenish what it metabolizes. By the time the flight modules reach the planet, we’ll have a good idea of how many generations it will take before we should schedule colonization. These ones are edible, too, which cuts down on setup costs.

3// The products in our branch typically aren’t edible, since certain features for survivability and military application use proteins that are toxic to humans. The few that are apparently taste terrible- it’s the same reason we don’t typically eat naturally occurring predators. Some products, like terraformers, we try to at least make them edible. It depends on the target atmosphere.

4// It’s not too uncommon for a module to become unusable during a campaign. Obviously it’s a huge profit and schedule loss, but it happens. Few actually die, but some results mean it doesn’t meet the requirement anymore and can’t be reworked to print. In those cases, we brick them and they get disposed of, or we keep them in stasis for display in the sales department. We did recently have one, I think it was a terraformer or secretion unit- something not for military application. And what happened was the engineer wrote the test procedure by copying and pasting a similar instruction that was used on a military program. Both programs required a ballistic survival test, but the caliber requirement was different. And the project engineer didn’t catch it, so the technician loaded a .50 caliber and blew the thing completely in half. I think it was a secretion unit because Todd was talking about the clean-up- how they needed insoluble spill kits for it.

5// All my bricks happen at vibration testing for some reason. Two were from the same thing- we typically hook units up to a pacemaker and simulate an alert state during the full-level random vibration to check for cardiac issues during launch or drop. Well, the design of the harness we use for the pacemaker is just flying leads that are pinched into place with thumb-screws. So on two different projects, the leads pulled out and shorted onto the chassis. First unit was a little guy, one of our standard oh-eighties. We didn’t even bother with a functional test, his heart just burst open. Second one was a big guy for a NASA program, we ended up arguing that his heart’s nominal cardioelectric pulse was greater than what the machine probably did to him. But that was a whole fiasco. Science missions are our worst critics by far, the military don’t give a shit about passing test failures.

6// Last week we got company hats that were- like five-panels? But they were orange and blue on different panels. It was to celebrate the last shipment of a big government program; the design required distinct chitin coloration so the units could be identified in the field. So we made a bunch of these acceptance units and they were partway through test when one of the engineers realized the colors were switched. They were supposed to be blue-over-orange but we’d made them orange-over-blue. So it was a whole big thing, and the customer wouldn’t budge so we had to brick, like, thirty of the things and start over from incubation. We eat the cost in those cases, they don’t pay us any extra.

7// Right, so imagine you have a robot. And the robot has circuits to indicate when a part of it is malfunctioning. It has a control board with code written to respond certain ways to certain malfunctions. Now, between the malfunction circuits and the control board, imagine instead of drawn copper conductor you have vat-grown synaptic fibers. That’s essentially what we’re doing with nerve grafting. Our products feel “pain” in the sense that they have conductors which communicate malfunction to a control system. But there’s nothing bad or good about that. It’s the collection and transfer of information. Whether the actuators are synthetic tissue or the control board is hand-spun synthetic axons makes no difference.

8// The specific processes our site uses are proprietary, but the basic practices are public domain, like we use in agriculture. A couple of companies offer subscriptions to sequence libraries, and Eden owns a handful of licenses. R&D engineers will select desired traits then crunch them with computer algorithms until they have a sequence that is likely to survive incubation. From there, the DNA is built via distillation in an amino processor and inserted into a blank synthetic cell. If the prototype survives long enough to demonstrate the intended qualities, the sequence can be patented.

9// Brains can be selected for in sequencing, and typical design philosophy is to overshoot the requirement and then condition it during manufacturing. In school you get a handbook with standard charts of mass-cognition ratios. So, if you need a product capable understanding simple commands, you might target a mass known to be capable of complex vocal patterns and then stunt it. Usually that’s done in incubation, but in some cases- like if we want to offer two variants- we apply the change later in manufacturing. Conk em with a hammer! No, I’m kidding. Typically they’re given a low-dose toxic inhalant or deprived of oxygen. For some of the really hardy military models, we lobotomize them. It sounds backwards, but it’s actually much cheaper to make many leaders and gimp all but one than it is to design and build two models.

10// Functional tests are performed with bioelectric harnessing that lets us convert nerve impulses into analog voltage signals. Some of the more complicated ones are performed with the unit in a CT scanner. Typically we look for malfunction signals in major organs, check isolation between nerve branches, measure stimulus response, and manually dilate major sphincters and valves. We perform functionals before and after any major environmental test to make sure nothing was broken. Usually it’s clear during the environmental testing if something’s gone wrong, but not always.

11// Capability tests are also normally done between environmental tests. We awaken the unit and observe for complex defects that wouldn’t show up in a functional test. Higher-order stuff like recognizing itself in a mirror, self-preservation instinct, command response. I had a unit once that came out of thermovacuum testing and the functional was fine, but when he was awake he would keep one arm up above his head unless he was using it for something. We wrote up a non-conformance and informed the customer, and engineering determined it wasn’t a failure to the specification. Didn’t affect form, fit or function. So we dispositioned him to use-as-is and shipped. He’s out there now, somewhere, wavin’ hi to everybody.

12// This is Mira, she’ll get you set up for payroll and whatever. I will probably see you tomorrow and we’ll get you familiar with the programs you’ll be supporting. Nice to meet you.