"Heroic Engineering"

[rust]

An important part of our campaigns is the engineers' and scientists' ability
to design and build improved or new equipment for personal use or com-
mercial production - we call this "heroic engineering".

Mongoose Traveller's Central Supply Catalogue has some rules for proto-
types and thelike, but these are quite basic, and for our next campaign I
would very much prefer a more detailed system.

I think I will start each project with a Concept Roll that will be at least Dif-
ficult for improvements and Very Difficult for new inventions. In my view
this first step of the project should take at least 1 week per attempt, with
a failed Concept Roll leading to no result and one with an Effect of -6 or
worse to design that is faulty and will lead to serious problems later on,
but without the character knowing this.

The next step of the project would be the Prototype Roll, again Difficult for
an improvement and Very Difficult for a new invention. The time required
to build the prototype probably depends on the size and complexity of the
device, but should be at least 2 weeks (testing included), and the price of
the prototype about 5 x the price of the finished device (to be determined
by the referee, I think). Again, a failed roll would result in a waste of time,
an Effect of -6 or worse in a future desaster.

The final step of the project would then be the Production. Since there is
already a working prototype, this should only require a normal skill roll,
parts with a cost of 50 % (?) of the finished device and half the time that
was needed to create the prototype.

Well, this is where I am currently - any comments, ideas and proposals
would be most welcome.

 

[kristof65]

The final step of the project would then be the Production. Since there is already a working prototype, this should only require a normal skill roll, parts with a cost of 50 % (?) of the finished device and half the time that was needed to create the prototype.

Production costs are generally based on volume. Prototypes aren't expensive just because of trial and error, but because they're generally only produced a few at a time - often by different means, IE, a prototype metal cabinet will be manufactured by hand, while a production run will be done on an automated line. Setup time of the production equipment is a huge factor that costs time and money.

With that in mind, the cost of a production unit should be dependant upon the volume run

 

[rust]

With that in mind, the cost of a production unit should be dependant upon the volume run

Thank you, a good point - I have to think about a way to enable the cha-
racters to make at least an educated guess concerning the cost of a pro-
duction line and the cost (parts plus labour) of the finished product, de-
pending on the number produced.

GURPS uses 20 times the price of the finished product as the cost of the
production line, but this still leaves the problem to determine the "base
price" of the product and the reduction of this base price by the number
of devices produced ... :?

 

[BP]

Volume as well as nature of product would have a very big impact on production costs - facilities could cost billions of times the product 'cost' (take ICs for example). Material, fabrication, assembly and packaging costs, as stated would also depend on volume. 'Cost', if modeled after U.S. consumer retail would be anywhere from 300%~1200% for non-niche items (which could go a lot higher - furniture, luxuries, jewelry, ...).

So, for selling price 'cost' I'd do 2d6x100% of 'production' cost, with DMs for market type.

For Traveller, of course, one has to make TL allowances..

 

[rust]

So, for selling price 'cost' I'd do 2d6x100% of 'production' cost, with DMs for market type.

Thank you, looks good, I will use it.

 

[Dave Chase]

Cost for prototypes are also very high if the items (parts) all have to be custom made.

If all the parts for the prototype are off the shelf then the cost of the prototype is still high but not as high as each part having to be made.

Dave Chase

 

[rust]

Well, it seems that any attempt to design a simple system for this is
bound to fail because there are too many variables, which would ma-
ke any remotely plausible system a lot more complex than I would
like, so I will have to decide these things on a case by case base.