Autonomous Starships

[agorski]

I'm trying to confirm what I'm reading in the T5 rules. Correct me if I'm wrong.

A ship's computer with a brain can control a number of functions simultaneously equal to the number of the computer (BIS +1).

Consoles are required to connect to control panels. A computer cannot connect directly to a control panel.

Consoles can monitor and adjust systems they control without a user or a brain, up to a point, like a life support system.

Consoles require a user, or a brain, or connection to a computer with a brain to control a system actively, like maneuvering.

Autonomous control could entail all consoles with brains and a computer without a brain, or a computer with brain and some or all consoles with brains, or a large computer with brain and no consoles with brains.

Consoles are sized and costed assuming a sapient user, so they're overkill for an autonomous setup. Pick a cramped one to minimize this.

Include a maintenance robot, workshop, small cargo hold, and tonnage for access passageways. For docked maintenance, consider adding some life support and add an air lock (personnel, cargo, or both).

 

[AnotherDilbert]

I'll give it a try...

A ship's computer with a brain can control a number of functions simultaneously equal to the number of the computer (BIS +1).

A computer does not need a brain to resolve tasks, as per T5.10:

A Computer resolves a Task (without a User) through a Console. Model Number is the number of simultaneous Tasks (Consoles) it may resolve.
A Brain is a permanently installed User-replacement in a Console.

Consoles are required to connect to control panels. A computer cannot connect directly to a control panel.

Normally, yes. Note that as per T5.10 a control panel need not be connected to a specific console.

Users can resolve tasks directly through a control panel (at added difficulty); presumably computers can too.

Consoles can monitor and adjust systems they control without a user or a brain, up to a point, like a life support system.

I think so. It's when you need to resolve a task that you need a user, computer, or brain.

Consoles require a user, or a brain, or connection to a computer with a brain to control a system actively, like maneuvering.

Yes.

Autonomous control could entail all consoles with brains and a computer without a brain, or a computer with brain and some or all consoles with brains, or a large computer with brain and no consoles with brains.

Enough computers would suffice, I believe. A bank of m/0bis computers can resolve tasks fairly cheaply.

Consoles are sized and costed assuming a sapient user, so they're overkill for an autonomous setup. Pick a cramped one to minimize this.

That would be reasonable, but you would have to house-rule Ergonomics.

Include a maintenance robot, workshop, small cargo hold, and tonnage for access passageways. For docked maintenance, consider adding some life support and add an air lock (personnel, cargo, or both).

Sure, but I think add more engineering robots; e.g. consoles break down all the time.

 

[agorski]

I would think that a computer without a brain can only handle simple control of mechanisms through a console. Like a console without a brain. Anyway, even if the rules allow it, I'd not want to trust an autonomous starship to a computer without a brain.

Users can resolve tasks directly through a control panel (at added difficulty); presumably computers can too.

I think this presumes that a user can poke at the buttons and blinky lights on a panel. A computer needs to go through a console to access the panel.

you would have to house-rule Ergonomics

How would ergonomics even apply with no crew aboard?

 

[AnotherDilbert]

I would think that a computer without a brain can only handle simple control of mechanisms through a console.

Why? An electronic brain is just a computer, albeit specialised?

How would ergonomics even apply with no crew aboard?

Presumably all the console break-downs are not just the crew bumbling around breaking things?

 

[agorski]

Why? An electronic brain is just a computer, albeit specialised?

Bk 2, pg 262:

Positronic Brain
Includes capabilities for Creative thought, free will, And the ability to learn

I think a brain is more sophisticated than a computer.

 

[AnotherDilbert]

Bk 2, pg 262:

That is an example of a typical robot, not a rule, note the source:

Robotics 101. Introduction to Robotics. Instructional Handout. University of Regina, Regina. 114-1104.

It says nothing about e.g. electronic brains. No matter how much we like Asimov, positronic brains are not the only possible.


A console can be fitted with any type of brain, without any functional difference to decision-making, see e.g. p249. An electronic brain is a (small) computer as far as I can see.

Electronic (or Photonic or Fluidic) Brain. A structure of electronic (less often fluidic or photonic) circuits operate as a logical data processor.

Note that computers can run Personalities, see p254.


I don't see any separation in capability, just operating principle, between the different types of brains, see e.g.:

Self-Aware missiles are equipped with a Brain (Electronic, Positronic, Semi-Organic, usually not Organic) which operates the missile and guides it to its target.

 

[whartung]

Bk 2, pg 262:

I think a brain is more sophisticated than a computer.

The difference between a "brain" and a "computer", if it's pinned to anything, I would put at "creativity".

Computer are anything but creative.

A computer can do everything a starship needs to travel. It can put the ship in orbit, run it to 100D, perform the Jump calc, press the big red JUMP button, decelerate out of Jump, and land the ship.

That's just rote work.

What it can't do is necessarily cope with the unknown.

 

[agorski]

So it sounds like a bunch of dumb consoles and a computer with a bunch of cells and a brain and maybe an automatic carousel of wafers would handle a starship just fine.