Gone are the days of sequestered insanity in darkened laboratories. Come, my fellow engineers, come into the light and bathe in the splendor of shared knowledge! Before I begin explaining all this, let's get some lexicon out of the way.
Prototype: A prototype is any device that you are still actively working on. Any good engineer will spend at least an hour a day working with any prototype he or she has, trying to perfect it. Devices in this stage are open to radical design adjustements and reconsiderations.
Schematic: Though I am loathe to admit, sometimes you must put your prototype down. One's attention cannot be divided amongst too many prototypes, or perhaps, more simply, you have lost interest in the device. By taking notes on what you have done so far, you enable yourself or another engineer to take up the project again later.
Production: A production model is a device that has been completed. Its design should not be altered, but it can be mass produced safely at this point, and sold to others, so they can enjoy the fruit of your labors. It is strongly recommended that only production models leave your workshop or university, as prototypes are notoriously prone to failure, and you can get a bad reputation allowing unfinished work to be sold.
Blueprints: A blueprint is a design note for a production model. If you create a design that would be excellent as a sub component of another device, selling the blueprints is the way to go, allowing other inventors to use your work in their own, or simply to create more of them.
Part: Any production design that is used in a new prototype is a part.
Disaster: Any prototype design that is used in a new prototype. Don't do this. You've been warned.
Invention: A combination of parts and/or disasters that has been brought together to make a more complicated device.
Make-Plan: So how does Make-Plan work in all this? For one, you can only have as many prototypes as your intelligence + Make-Plan, divided by 10. If you had max in both(50 total), you could have 5 prototypes. You can always have 1 prototype at minimum, so long as you have at least 1 point in make-plan. Some designs require multiple prototypes to complete, and thus require a shrewd and masterful inventor to pull off, managing these different disasters (See lexicon above) in one great juggling act.
Power Sources -
All great inventions need a power source. There are a few exceptions. Medicines and poisons derive their 'power' from the chemicals that also create the action desired, for example. These are rare inventions, however. If you do not already have a blueprint or disaster to use as a power source, you must make one before you can get to work on the design. Power systems are very slow to develop, but, often, can be used and re-used as parts in many inventions to come. A roll can be made once per two and a half weeks (1 OOC week). The roll is Intelligence+Science/(Power Type)+Make-Plan vs 16
Controlling Devices -
Unless the device is 'turned on' in process of completing it and there's no reason to ever redirect or turn it off, it will need a controlling device, the second critical part to any complicated device. Controls are not as difficult as power sources, Allowing three rolls every OOC week. The roll is Intelligence+Science/(Control Type)+Make-Plan vs 16
Action Devices -
What use is a device that doesn't do anything? This is the exciting part for any engineer. Action parts, designed with controlling and power parts, are what create new scientific marvels for the world to enjoy! The roll for action parts is Intelligence+Make-Plan vs 10, though the time required for each roll varies with each part. Some also have other requirements.
Fine Tuning: All prototypes can be fine tuned on demand, making them perform better, or differently. Doing so makes the device less stable, but can be very handy. Make a Dexterity+Intelligence+Fine-Tune roll. Add +2 successes per heretic tier you have. Use this resulting number to add or subtract from the ratings of the target prototype. It must be your prototype. Every 2 points subtracted is one more point that can be added somewhere else. These changes will revert at the end of the scene. Using this ability costs 3d20 gil in spare parts. Every time the device is turned on or used or the benefits are employed, there is a 10% chance the item fails to operate while fine tuned. If the die roll comes up 1, the device fails spectacularly, and will need repairs in addition to possibly exploding, or some other equally dramatic event.
Finalization: So you're happy with the invention and you want to push it out of prototype and into final production model? Easier said then done! First you must cease improving the invention. You must then spend as much XP as 1/10th the total points involved in the invention. You do not have to repay the points in already invented parts that make up the invention. The minimum cost to finalize is 2 XP. Once finalized, the invention ceases to be prototype and is now production. You are also assumed to have a blueprint, which you can give/sell to others.
If the item contains disasters, this is where things get tricky. You must pay the XP cost for all disasters and the final object at once. There is a 5% chance per disaster that the entire thing fails, losing your work (You don't spend XP if this happens, and roll each disaster separately). If the invention fails, there is a separate 5% per disaster chance that the device fails spectacularly, possibly causing harm to yourself and/or your laboratory. Inventions that are lost at this step are completely lost. A sad day.
There is a financial investment as well, consisting of 1 gil per point involved, INCLUDING already invented parts that make up the final invention.