Self-repairing robots

[Rivenshield]

On page 63 of 3e Robots there is a Utility program called Repair. At $500 and Complexity 1 at TL8 it is eminently affordable, and gives the robot with the right tools 'a skill of (program TL+4) for repairs and routine maintenance only, for one specific gadget.'

This is a heck of a lot cheaper and more effectual than paying out the nose for a Mechanic skill program at $2000 per. point for enough points to have a decent chance of success and a high-complexity robot brain to run it, if all you want to repair is one gadget. Of course I understand that 'gadget' generally refers to the stuff in Ultratech -- datapads, laser pistol, vapor canteens, etc. -- but could it be construed to mean larger more complex devices, as long as it was a single robot or roboticized vehicle?

After all, why *wouldn't* a futuristic car be self-tuning? Once a month several retractile arms whip out of hidden hatches, the hood opens, it performs a routine 1000-point checkup, and calls the factory for any new parts required. It couldn't repair your automated lawnmower, and would choke if you installed any aftermarket parts.... but as long as your car is factory stock, you're fine. What about the military applications? Wouldn't a self-maintaining battlesuit solve a lot of logistical problems? A self-repairing tank? Combined with Damage Control software it could put *itself* back into action, as long as you supplied the spare parts.

What are your thoughts, o gearheads? And what would you allow in your own campaign, o gamemasters? I want to conform to the rules as written, not do an end-run around them, but the premise seems reasonable enough.

[RyanW]

It doesn't seem out of line for self-diagnosis, but actual repair of something as complex as a complete vehicle feels like too much to me.

[Anaraxes]

The program is just information. Actually carrying out the repairs needs tools, parts, and hands (of some sort). A fully self-repairing gadget would need those tools, parts, and manipulators. A vehicle or battlesuit probably doesn't have extra payload capacity for one of everything they might need. So, they'd need access to a depot; that depot would have to be designed for them (so the tank can drive into the warehouse and retrieve parts, so probably not your typical hardware store for humans layout). And so on. That tank is going to need some sort of extensible arms, service drones, whatever. All that stuff is extra cost, weight, and points of failure that need to be taken into account when assigning vehicle stats, on top of just the robot brain and program. (Might well be wiser to make the repair depot the gadget with brains and remote waldo bots, instead of duplicating it all in each individual gadget.)

But given all that -- sure, why not build self-repairing devices? If they're sufficiently complex (starships, say, if not tanks), you might well up the price and complexity requirements of the Repair program, as it's certainly more involved than repairing a datapad.

[WingedKagouti]

Originally Posted by Anaraxes

But given all that -- sure, why not build self-repairing devices? If they're sufficiently complex (starships, say, if not tanks), you might well up the price and complexity requirements of the Repair program, as it's certainly more involved than repairing a datapad.

You also need maintenance on the self-maintenance kit along with materials to aid in all repairs in case materials can't be located in the environment.

Also, while it's not completely impossible to make manipulators that can reach every part, it's easier to make two symbiotic machines that can repair each other.

[Anthony]

Originally Posted by Anaraxes

But given all that -- sure, why not build self-repairing devices?

Mostly because making devices possible to repair often makes them less reliable. There's a reason a lot of modern devices are designed on a principle of "if it stops working, get a new one", and it's not just the manufacturer's desire for profit.

On the repair program: it has skill 12 at TL 8, which means we can compare it to real world devices. Now, real world devices can certainly diagnose a fair number of problems for themselves, but they're all on the scale of "adequate to let an untrained person push the right button". I would probably delete the +4, a skill of 8 is perfectly fine for trivial stuff that will have a bonus.

[Pursuivant]

Originally Posted by RyanW

It doesn't seem out of line for self-diagnosis, but actual repair of something as complex as a complete vehicle feels like too much to me.

It depends.

If "self-repair" depends on simply bypassing damaged but redundant systems or tweaking sensors which regulate fuel or energy flow it could be realistic even with current technology.

Something which isn't listed/mentioned is automatic damage control software, which should cost about as self-diagnosis software, but with additional cost, volume, and mass for the automated DC systems. This sort of system would give Mechanic or Crewman skill roll at some fixed cost, but only for purposes of preventing further damage and limiting problems like flooding, fire, or fuel loss.

[Anaraxes]

Originally Posted by Anthony

Mostly because making devices possible to repair often makes them less reliable.

That's a perfectly good real-world reason. Cost and complexity are going to be much higher for a self-repairing device, due to the wide variety of extra requirements that would otherwise be unnecessary. That's more things that can go wrong, as well as more cost, weight, etc. Short of reconfiguring universal nanotech magic, that is...

With that remark, though, I was really answering the question in the OP -- more of a rules/gamist-based question about why the Repair program is cheaper than the Mechanic program, and whether self-repairing gadgets ought to be allowed in a campaign. (As always with setting details, the answer is "how do you want it to work". "Are the rules going to break or get exploited" is worth considering, but it really helps to keep the desired end goal in mind while you're navigating all the rules and options.)

[Daigoro]

At Complexity 1, it sounds like it's just a collection of blueprints and a maintenance manual for the device, with whatever level of detail and rendering is appropriate for the TL. If that allows a robot to do repairs, it sounds like the robot already has that general capability and this software just gives it the knowledge to do it properly on that single device.